Professional Documents
Culture Documents
Automatic Control Systems by Benjamin C Kuo PDF
Automatic Control Systems by Benjamin C Kuo PDF
BENJAMIN
C.
KUO
utomatic
control
THIRD EDITION
HI
OX
2D
-'.
PRLNIlUt
HALL
Syste
Automatic
Control
Systems
Third Edition
BENJAMIN
C.
KUO
PRENTICE-HALL,
of Illinois at Urbana-Champaign
INC., Englewood
Cliffs,
New Jersey
(,^-s
in Publication
Kuo, Benjamin
Data
Automatic control.
2.
Control theory.
Title.
TJ213.K8354 1975
ISBN 0-13-054973-8
629.8'3
74-26544
METROPOLITAN
BOROUGH OF W1GAN
OF
DEPT.
LEISURE
LIBRARIES
*?
10067*2
A-
Ace. No,
o
:
5m^?1^;,
1975 by Prentice-Hall, Inc.
Englewood
Cliffs,
New Jersey
No
may be reproduced
in
without permission in
part of this
book
10 9 8 7 6 5 4
PRENTICE-HALL
PRENTICE-HALL
PRENTICE-HALL
PRENTICE-HALL
PRENTICE-HALL
OF
OF
OF
OF
CANADA,
LTD., Toronto
JAPAN,
INC.,
Tokyo
New
Delhi
Contents
Preface
Introduction
1.1
Control Systems
1.2
What
1.3
2.
Mathematical Foundation
15
15
2. 7
Introduction
2.2
Complex-Variable Concept
2.3
Laplace Transform
15
18
2.4
2.5
2.6
2.7
Matrix Algebra
2.8
z-Transform
Differential Equations
21
25
26
32
39
~\
vi /
3.
Contents
3.2
51
3.3
55
3.4
Block Diagrams
3.5
3.6
Summary
3.7
3.9
3.
4.
10
58
64
Flow Graphs
69
Signal-Flow-Graph Algebra
Flow Graphs
67
66
71
3.11
3.12
Systems
80
81
95
4.
4.2
State Equations
4.3
4.4
4.5
4.6
4.7
4.8
4.9
97
99
101
103
Relationship
Form
109
Between State Equations and Transfer Functions
117
Characteristic Equation, Eigenvalues, and Eigenvectors
Transformation to Phase-Variable Canonical
Relationship
4.10
Diagonalization of the
4.11
4.14
4.15
Controllability of Linear
4.16
4.17
Relationship
12
4.
13
4.
18
Among
Systems
144
152
Controllability, Observability,
Transfer Functions
156
158
161
4.20
165
4.21
State
4.23
167
Diagram for Discrete-Data Systems
171
State Diagrams for Samp/ed-Data Systems
State Equations of Linear Time-Varying Systems
173
5.
Introduction
5.2
5.3
5.5
5.6
Tachometers
5.4
118
State
4.19
4.22
115
123
Diagram
126
136
Decomposition of Transfer Functions
141
Transformation into Modal Form
4.
and
5.
51
3.1
3.8
51
219
188
190
208
187
Contents
5.7
5.8
5.9
5.10
5.11
6.
5.
12
5.
13
259
Introduction
6.2
6.5
6.6
6.7
6.9
6.
7.
10
295
Stability of Control
302
305
Systems
316
316
7.1
Introduction
7.2
Stability. Characteristic
7.3
Equation,
7.4
Methods of Determining
7.5
Routh-Hurwitz
Criterion
317
Transition Matrix
319
Systems
321
322
330
7.6
Nyquist Criterion
7.7
7.8
7.9
8.
259
6.
6.4
344
7.10
7.11
Stability of
the
Shape
375
375
8.
Introduction
8.2
8.3
8.4
376
380
412
8.5
Some
8.6
8.7
8.8
8.9
vii
DC
6.3
444
417
viii /
9.
Contents
459
9.
Introduction
9.2
9.3
462
Frequency-Domain Characteristics
Second-Order System
p .CO p , and the Bandwidth of a
9.4
9.5
9.6
9.7
Relative Stability
Open-Loop
Open-Loop
464
Transfer Function
467
Transfer Function
471
and
9.9
9.10
Constant
Mp
473
9.8
10.
12
Versus-Phase Plane
496
504
Introduction
10.2
Systems
10.3
Phase-Lead Compensation
Phase-Lag Compensation
515
535
552
10.4
10.5
10.6
11.
490
504
510
Lag-Lead Compensation
Bridged-T Network Compensation
557
572
11.1
Introduction
1 1.2
Analytical Design
574
583
1.3
Parameter Optimization
1.4
1.5
11.6
588
Design of State Observers
Optimal Linear Regulator Design
11.7
Design with
Partial State
Feedback
599
615
APPENDIX A Frequency-Domain
Plots
627
A.
A.2
A.3
Bode
Magnitude-Versus-Phase Plot
626
633
643
APPENDIX B
APPENDIX C
Lagrange's Multiplier
Index
585
Method
645
650
653
Preface
The
The two major changes in the second edition, published in 1967, were the
inclusion of the state variable technique and the integration
of the discrete-data
systems with the continuous data system. The chapter on
nonlinear systems
was eliminated in the second edition to the disappointment of some
users of
that text. At the time of the revision the author felt that
a comprehensive treatment on the subject of nonlinear systems could not be made effectively
with
The third edition is still written as an introductory text for a senior course
on control systems. Although a great deal has happened in the area
of modern
control theory in the past ten years, preparing suitable material
for a
course on introductory control systems remains a difficult task.
it is difficult
to
modern
The problem is
teach the topics concerned with new
the undergraduate level. The unique
modern
Preface
emphasis
and intermediate levels. Therefore in this current edition, equal
theory.
control
modern
the
and
methods
classical
placed on the
A number of introductory books with titles involving modern control
is
have attempted to
theory have been published in recent years. Some authors
but according
control,
modern
the
with
control
classical
the
integrate
unify and
is highly
goal
such
a
Although
failed.
have
most
reviews,
and
the critics
to
desirable, if only
be a good
new
solution. It
theories
remains that
is
may
to
and new techniques are developed for this purpose. The fact
of
control systems, in some way, may be regarded as a science
how
to solve one
method and the modern approach indepenconsidered as alternadently, and whenever possible, the two approaches are
are weighed. Many
tives, and the advantages and disadvantages of each
this text is to present the classical
illustrative
Many
many
text
book
that the
experience necessary to provide real-life examples. Another reason is
realmost
fact
that
the
difficulty in the control systems area is compounded by
problems are highly complex, and are rarely suitable as illustrative examples
is lost by simplifying
at the introductory level. Usually, much of the realism
developed in the
techniques
the problem to fit the nice theorems and design
taking
a control system
students
material. Nevertheless, the majority of the
life
text
must put
course at the senior level do not pursue a graduate career, and they
extremely
is
It
employment.
new
their knowledge to immediate use in their
an
important for these students, as well as those who will continue, to gain
life
problems.
The following
with the
first
1.
2.
3.
features of this
new
two editions
in this
book
is
an outgrowth of a
senior-level
Illinois at
control system course taught by the author at the University of
in a style
written
Urbana-Champaign for many years. Moreover, this book is
Preface
/ xi
Chapter
3.
Chapter 4 intro-
These subjects are later being used for the analysis and
design of linear control
systems. Chapter 5 discusses the mathematical
modeling of physical systems.
Here, the emphasis is on electromechanical systems.
Typical
transducers and
control systems used in practice are illustrated.
The treatment cannot be
exhaustive as there are numerous types of devices and
control systems. Chapter
6 gives the time response considerations of control systems. Both the classical
and the modern approach are used. Some simple design
considerations in the
time domain are pointed out. Chapters 7, 8, and 9 deal
with topics on stability,
root locus, and frequency response of control systems.
In Chapter 10, the design of control systems is discussed,
and the
approach
is
basically classical.
Chapter
11 contains
subjects
One of
subjects to cover.
To keep
book
draft, had
the
to a reasonable length,
Benjamin C.
Urbana,
Illinois
Kuo
1
Introduction
1 .1
Control Systems
living. Industrially,
The
1-1.
The
may be
signal e
is
control system.
In more common terms, the controlled variable is the output of the system,
and the actuating signal is the input. As a simple example, in the steering control
of an automobile, the direction of the two front wheels may be regarded
as the
controlled variable
c,
e.
the output.
However,
Chap.
Introduction
Controlled
Actuating
signal e
variable c
Control
system
(Input)
(Output)
There are many situations where several variables are to be controlled simulmultivariabk
taneously by a number of inputs. Such systems are referred to as
systems.
output variable back and comparing it with the command signal. When a system
does not have the feedback structure, it is called an open-loop system, which is
the simplest and most economical type of control system. Unfortunately, openloop control systems lack accuracy and versatility and can be used in none but
the simplest types of applications.
Consider, for example, control of the furnace for
home
heating. Let us
assume that the furnace is equipped only with a timing device, which controls
the on and off periods of the furnace. To regulate the temperature to the
proper level, the human operator must estimate the amount of time required
for the furnace to stay on and then set the timer accordingly. When the preset
time is up, the furnace is turned off. However, it is quite likely that the house
temperature is either above or below the desired value, owing to inaccuracy in
the estimate. Without further deliberation,
control
is
one
fact that
may
not
know
One
it is
lies
in the
The other
is
is
operating period, the final temperature inside the house will not be accurately
regulated by the open-loop control.
An
electric
washing machine
is
A
and estimation of the human
machine should have the means of checking the cleanliness of the clothes continuously and turn itself off when the desired degree of cleanliness is reached.
Although open-loop control systems are of limited use, they form the basic
Sec. 1.1
Control Systems / 3
1-2. An input
applied to the controller, whose output acts as the
actuating signal e; the actuating signal then actuates the controlled
process and
hopefully will drive the controlled variable c to the desired value.
signal or
command
r is
Reference
Actuating
input r
signal e
Controller
Controlled
variable c
Controlled
process
(Output)
Fig. 1-2.
What is missing in the open-loop control system for more accurate and
more adaptable control is a link or feedback from the output to the input of
the
system. In order to obtain more accurate control, the controlled
signal c(t) must
be fed back and compared with the reference input, and an
actuating signal
proportional to the difference of the output and the input must be
sent through
the system to correct the error.
system with one or more feedback paths like
that just described is called a closed-loop system.
Human beings are probably
the most complex and sophisticated feedback control system
in existence.
human
being
may
To
illustrate the
many
inputs and
human
is
to reach for
may be
is
as an open-loop system.
being
is
As another
illustrative
human
Fig. 1-4
Error
Input
detector
command
f x
Error
Controller
(brain)
Reach
for object
Fig. 1-3.
Controlled
process
1 Controlled
variable
Position
of hand
Chap.
/ Introduction
Rudder
Fig. 1-4.
Rudder
control system.
shows the block diagram of the rudder control system of a ship. In this case the
objective of control is the position of the rudder, and the reference input is
applied through the steering wheel. The error between the relative positions of
the steering wheel and the rudder is the signal, which actuates the controller
When
the rudder
is finally
is
zero. Let us
wheel position
is
have
electrical
the position of the rudder cannot respond instantaneously to a step input, but
will, rather, move gradually toward the final desired position. Often, the response
will oscillate
about the
rudder control
it is
final position
before settling.
It is
0,(0
6e
*~t
-*-t
(a)
(b)
Fig. 1-5. (a) Step displacement input of rudder control system, (b) Typical
output responses.
Sec. 1.1
Control Systems / 5
Error
sensor
Input
~^
Error
Controller
Controlled
process
Output
Feedback
elements
are
The basic elements and the block diagram of a closed-loop control system
shown in Fig. 1-6. In general, the configuration of a feedback control system
may not be constrained to that of Fig. 1-6. In complex systems there may be a
multitude of feedback loops and element blocks.
Figure l-7(a) illustrates the elements of a tension control system of a windup
process.
which
is
The unwind
collects it by winding it onto another roll. The control system in this case is
to
maintain the tension of the material or web at a certain prescribed tension to
avoid such problems as tearing, stretching, or creasing.
roller.
the
unwind
reel is to generate
horizontal position at
all
times.
To
is
Chap.
Introduction
Unwind
reel
Web
(decreasing dia.)
processing
Windup
reel
(increasing dia.)
Drive system
(constant
web
speed)
(Current)
Reference
input
~"\ Error
Controller
Electric
Unwind
brake
process
Tension
Dancer
arm
(b)
1 .2
What
Is
Its
Effects ?
control systems.
We demon-
closed-loop
strated in Section 1.1 that feedback is a major requirement of a
able
to achieve
not
be
would
system
control
control system. Without feedback, a
applications.
practical
in
most
required
are
the accuracy and reliability that
However, from a more rigorous standpoint, the definition and the significance
feedback are much deeper and more difficult to demonstrate than the few
of
carry
examples given in Section 1.1. In reality, the reasons for using feedback
the
output
with
input
the
comparing
one
of
simple
the
than
far more meaning
error is merely one of the
in order to reduce the error. The reduction of system
We shall now show that
system.
upon
a
bring
may
feedback
that
effects
many
Sec.
.2
What
Is
Its
Effects ? /
stability,
there are
when
with or
without physical feedback, to be studied in a systematic way once the existence
of feedback in the above-mentioned sense is established.
show
is
G
M = t = FTW
Using
some of the
(l-i)
significant effects
_.
i
+
r
-o
b +
H
-o
Fig. 1-8.
Feedback system.
As seen from Eq. (1-1), feedback affects the gain G of a nonfeedback system
by a factor of 1 + GH. The reference of the feedback in the system of Fig. 1-8
is negative, since a minus sign is assigned to the feedback
signal. The quantity
GH may itself include a minus sign, so the general effect of feedback is that it
may
and
H are
Chap.
Introduction
+ GH
it
in another.
=-
finite input.
Therefore,
originally stable to
GH =
is
+GH+GF
( "
and
GH =
H are
1,
+-
o+
c
-o
o-
-o
H
-o
o-
o-
-o
Fig. 1-9.
all
an important
Sec
1-2
What
Is
Its
Effects? / 9
system to be completely stationary over the entire operating life of the system.
For instance, the winding resistance of an electric motor changes as the temperature of the motor rises during operation. In general, a good control system
command responsively. We
sensitivity to
The
effect
still
able to follow
parameter variations.
what
shall investigate
sensitivity
we
consider
G as
may
a parameter that
M to the variation in G
is
defined as
_~ dM/M
io
^- 3 >
~dGjG
in
dM G _
SM
io _
~lGM~l+GH
>
This relation shows that the sensitivity function can be made arbitrarily small
by increasing GH, provided that the system remains stable. It is apparent that
in an open-loop system the gain of the system will respond in a one-to-one
fashion to the variation in G.
In general, the sensitivity of the system gain of a feedback system to paramdepends on where the parameter is located. The reader may
derive the sensitivity of the system in Fig. 1-8 due to the variation of H.
eter variations
in electronic amplifiers
The
effect
duced into the system; no general conclusions can be made. However, in many
can reduce the effect of noise on system performance.
situations, feedback
Let us refer to the system shown in Fig. 1-10, in which r denotes the comsignal and n is the noise signal. In the absence of feedback,
0, the
output c is
H=
mand
where e
r.
The
=GGe+G
x
2n
(1-5)
_
GGe_
c
~~
x
defined as
is
G2 n
'
~n
'
no
effect
whatsoever on the
ratio.
to r
and n acting
10
Chap.
Introduction
\h
+
+
b
G2
e2
Gi
__
_.
Fig. 1-10.
simultaneously
is
Gl G 2
T + G,G 2 H
+ +
_|
b3
G,G 2 H
(1-7)
K
'
ponent
signal.
(1-7)
is
_GG
+ G^G^H
+ G G H)
2 rj(\
2 n/(l
___
r_
1
(1-%}
and is the same as that without feedback. In this case feedback is shown to have
no direct effect on the output signal-to-noise ratio of the system in Fig. 1-10.
However, the application of feedback suggests a possibility of improving the
signal-to-noise ratio under certain conditions. Let us assume that in the system
of Fig. 1-10, if the magnitude of G is increased to G\ and that of the input r
to r', with all other parameters unchanged, the output due to the input signal
t
acting alone
we
is
at the
same
level as that
when feedback
is
'1- ^
let
(1 ' 9)
which
is
to-noise ratio
is
is
when G
is
becomes
now
G 2 nl{\
which
G\G 2 H)
^ +
^^>
(1-11)
(1
G\G 2 H).
In general, feedback also has effects on such performance characteristics
Seo
Types
of
11
1.3
become known
as
control the output variables according to the way the names imply. In general,
there are many other ways of identifying control systems according to some
special features of the system. It
ways of
is
known
common
is
This classification
is
made according to
do not
and
design.
tems are nonlinear to some extent. Linear feedback control systems are idealized
models that are fabricated by the analyst purely for the simplicity of analysis
and design. When the magnitudes of the signals in a control system are limited
to a range in which system components exhibit linear characteristics (i.e., the
principle of superposition applies), the system is essentially linear. But when the
magnitudes of the signals are extended outside the range of the linear operation,
depending upon the severity of the nonlinearity, the system should no longer be
considered linear. For instance, amplifiers used in control systems often exhibit
saturation effect when their input signals become large; the magnetic field of
a motor usually has saturation properties. Other common nonlinear effects
found in control systems are the backlash or dead play between coupled gear
members, nonlinear characteristics in springs, nonlinear frictional force or torque between moving members, and so on. Quite often, nonlinear characteristics
are intentionally introduced in a control system to improve its performance or
provide more effective control. For instance, to achieve minimum-time control,
an on-off (bang-bang or relay) type of controller is used. This type of control is
found in many missile or spacecraft control systems. For instance, in the attitude
control of missiles and spacecraft, jets are mounted on the sides of the vehicle
to provide reaction torque for attitude control. These jets are often controlled
in a full-on or full-off fashion, so a fixed amount of air is applied from a given
jet for a certain
12
Chap.
Introduction
For
and graphical
tech-
niques for design and analysis purposes. However, nonlinear systems are very
that may be
difficult to treat mathematically, and there are no general methods
When
time during the operation of the system, we have a time-invariant system. Most
physical systems contain elements that drift or vary with time to some extent.
the
If the variation of parameter is significant during the period of operation,
unwind
the
of
system is termed a time-varying system. For instance, the radius
of the tension control system in Fig. 1-7 decreases with time as the material
being transferred to the windup reel. Although a time-varying system without
reel
is
nonlinearity
is still
a linear system,
its
analysis
is
usually
electrical engineering,
system is referred to, it does not mean that all the signals in the system are of
dc control
the direct-current type; then there would be no control movement.
system simply implies that the signals are unmodulated, but they are still ac by
common definition. The schematic diagram of a closed-loop dc control system
is
^^^
6*^) "r
Reference
input
shown
in response to a step
Error
detector
Controlled
variable
6,
diagram of a
typical
Sec. 1.3
13
sys-
in Fig. 1-12.
system
all
Synchro
transmitter
a-c
servomotor
Reference
input
0.
Fig. 1-12.
may be
Chap.
14
Introduction
Input
eg)
r(t)
*c >
Sampler
Data
hold
hit)
Controlled
process
c(f)
(filter)
Fig. 1-13.
sampler, and
is sampled by a sampling device, the
rate of the samsampling
The
pulses.
of
sequence
is
a
sampler
the output of the
advantages of incorporating
pler may or may not be uniform. There are many
understood of these being
easily
the
most
of
one
system,
control
sampling in a
equipment among several
expensive
of
an
sharing
time
provides
that sampling
system.
The
control channels.
flexibility,
Because digital computers provide many advantages in size and
airMany
computer control has become increasingly popular in recent years.
discrete
borne systems contain digital controllers that can pack several thousand
1-14 shows the
elements in a space no larger than the size of this book. Figure
Digital
coded
input
Digital-to-
Digital
computer
,
analog
converter
missile
Airframe
Analog-to-
con\ erter
Fig. 1-14. Digital autopilot system for a guided missile.
2
Mathematical Foundation
2.1
Introduction
The study of control systems relies to a great extent on the use of applied
mathematics.
For the study of classical control theory, the prerequisites include such
complex variable theory, differential equations, Laplace transform,
subjects as
2.2
Complex-Variable Concept
Complex-variable theory plays an important role in the analysis and design
of
When studying linear continuous-data systems, it is essential
that one understands the concept of complex variable and functions
of a complex
control systems.
variable
when
method
is
used.
Complex Variable
complex variable j is considered to have two components: a real component a, and an imaginary component co. Graphically, the real
component is
represented by an axis in the horizontal direction, and the imaginary
component
measured along a vertical axis, in the complex j-plane. In other words,
a complex variable is always defined by a point in a complex plane
that has'
a a axis and ayco axis. Figure 2-1 illustrates the complex j-plane,
in which any
is
15
Chap. 2
16
Mathematical Foundation
/co
s-plane
OJ]
i
Fig. 2-1.
arbitrary point, s
simply
Si
ffj
= su
is
Complex
j-plane.
and
== a,
co
a> or
+y'coi.
The function
G(s)
is
G(j)
where Re
part of
= ReG+yImC
(2-1)
measures
/co
S,
=0, +/C0,
ImG
G 0)-plane
ReG
a,
Gfri)
the G(-plane.
Sec 2 2
-
Complex-Variable Concept
17
G(J)=
<2
,-(7TT)
"
2)
Analytic Function
= +
Singularities
its
are the points in the j-plane at which the funcpole is the most common type of singu-
theory.
The
definition of a pole
order r at s
lim [0
- s,)
G(s)]
the factor (s
W=
l0(s
G(s)
= -3
is
s(s
+ 2)
+ 3)*
n
(2
IX*
" xi
3>
=-
at s
and s
1. It can
analytic in the j-plane except at these poles.
Zeros of a Function
The
is
definition of a zero of
analytic at s
=s
t,
it is
!S [(* ~
has a finite, nonzero value.
has an rth-order pole at s
Or
zero at s
s,.
J '>" ,<7
W]
( 2 -4)
2.
If the function
18
r.
Chap. 2
Mathematical Foundation
= lim^=0
S
limGO)
(2
"5
s-<*>
s-.
s-
plane.
2.3
3-5
Laplace Transform
is
2.
one operation.
differential equation into an
then possible to manipulate the algebraic
s.
It is
s
equation by simple algebraic rules to obtain the solution in the
Laplace
inverse
the
taking
by
obtained
is
solution
domain. The final
transform.
Definition of the Laplace Transform
satisfies
the condition
(2-6)
r\f(t)e-'\dt<oo
J
for
some
finite real a,
F(s)=
m=
or
The
variable s
variable; that
is,
is
is
defined as
\~ f(t)e-"dt
(2-7)
(2
[/(')]
a +jco. The
which
is
is
a complex
also
evaluated from
is
as the one-sided Laplace transform, as the integration
This simply means that all information contained in /(f) prior to
-g
is
known
to oo
ignored
may
be used
Sec. 2.3
Laplace Transform / 19
Example
2-1
Let/0) be a
of unity for
>
F(s)
< 0.
(2-9)
= [u (t)] =
s
j~ us )e-" dt
e~"
(2-10)
j
which means that the
ut)e-" dt
real part
Or,
is
Of course,
for
= u.(t)
/(/)
Then
is
e-
dt
is
<
valid if
co
we simply refer to the Laplace transform of the unit step function as lis, and rarely do
we have to be concerned about the region in which the transform integral converges
absolutely.
Example
2-2
where a
is
= e-",
2;
a constant.
is
written
=/: e--'e-" dt
+a
+a
(2-11)
The operation of obtaining /(?) from the Laplace transform F(s) is termed
the inverse Laplace transformation. The inverse Laplace transform of F(s) is
denoted by
f(t)
and
is
Z-Vis)]
(2-12)
/(0
2^7
'
me"ds
(2-13)
where
c is a real constant that is greater than the real parts of all the singularities
of F(s). Equation (2-13) represents a line integral that is to be evaluated in
the j-plane. However, for most engineering purposes the inverse Laplace transform operation can be accomplished simply by referring to the Laplace trans-
form
table,
The
by the
many
presented in the following in the form of theorems, and no proofs are given.
20
Chap. 2
Mathematical Foundation
by a Constant
The Laplace transform of the product of a constant k and a time function f{t) is the constant k multiplied by the Laplace transform of f{t);
Multiplication
that
is,
[kf(t)]
2.
= kF(s)
(2-14)
the
sum
tions; that
is,
[fi(0
F {s)
where
and
2 {s)
(2-1 5)
and/2 (r),
respectively.
3.
Differentiation
is
approaches
0-\-; that
is,
= sF(s) -
df(ty
lim /(/)
dt
(2-16)
sF(s)
/(0+)
[d'fW] _
df
s"F(s) -- lim
s-if(t)
s~
ldl^-\-
(-0 +
-- ^r /(o+)
n
s F(s)
5"" 2
(2-17)
4. Integration
is
first integral
S\
F{s)
f(j)dr
is,
(2-18)
...
fir) dx A,
rr...r.
J
Jo
J
5.
Shift in
dt n
(2-19)
Time
[fit
where u s {t
T)us {t
T)]
is,
= e- T *F{s)
(2-20)
is
shifted in
Sec 2 4
-
6.
Initial-Value
Theorem
is
F(s), then
lim sF(s)
(2-21)
Final-Value Theorem
7.
lim sFis)
(2-22)
The final-value theorem is a very useful relation in the analysis and design
of feedback control systems, since it gives the final value of a time function
by
determining the behavior of its Laplace transform as s tends to zero. However,
the final-value theorem
not valid
is
Example
2-3
is
Since sFis)
is
analytic
theorem
final-value
on the imaginary
may
2-4
+s+2)
axis
and
lim fit)
Example
s(s 2
5
,
,
,
= -1
f2-231
F& =
J2"^p
(2-24)
which is known to be the Laplace transform of /(f) = sin cot. Since the function
sFis)
has two poles on the imaginary axis, the final-value theorem cannot be
applied in this
case. In other words, although the final-value theorem would
yield a value of zero as
the final value of fit), the result
2.4
is
erroneous.
'
and
partial-fraction
expansion.
When the
function in
s, it
X^ =
W)
is
a rational
22
Chap 2
Mathematical Foundation
where P(s) and Q(s) are polynomials of s. It is assumed that the order of Q(s)
in s is greater than that of P(s). The polynomial Q(s) may be written
Q(s)
a^"'
+a
n-
(2-26)
where
Partial-Fraction Expansion
X( S)
(2-27)
t ,
written
X(vS')
The
coefficient,
(i
= -^+
+ -4*s + s
s + i,
1, 2,
n), is
K
j,
(j + Ji) and let s =
To
(S
s\
As an
illustrative
sl
that
The
is
(s
for instance,
+ T^V
s
s
<2
28 )
-1--
we
is,
+ s )?&
(s 2
Pis,
s t )(s 3
Si)...
(s
(2-29)
Si)
(s
which
l)(,s
2)(s
3)
coefficients
K_ u K_ 2 and K_ 3
*-.
*-2
+3
+ W*)],-i = ( 2-~lX3 - l) _i
~ 2) + 3 =
5(
7
= + 2)X(s)} ^ = (12X3 - 2)
5(--3) + 3
-6
= [(* + 3)*(j)],-, (13X2 - 3)
[('
[(.s
._
A"- 3
,-=
(2-32)
(2-33)
(2-34)
W) = tti + s +7 2 - t^
j + 3
2 - 35 >
Sec. 2.4
Partial-Fraction Expansion
If r of the
multiplicity
r,
X(s)
/>(*)
g(5)
Then
JT(j)
A*)
(5
= ^^
J + s
<
I
s t Xs
*.
**)...(*
*,)'(*
is
of
(2-36)
(2-37)
A,
(*
of simple poles
A2
AT,,
...+
r) terms
+s +
A,
s,
can be expanded as
X(s)
The n
written
is
*(*)
23
J,)'
>
|
is
described below.
= [(S + *,)'*(*)],__
Ar
(2-38)
Ar-^-Ks + s^Xis)]
(2-39)
(2-40)
(2-41)
Example
2-5
X{s)
s{s
(2-37),
*
J
Then
X(s)
l)3(j
+ 2)
(2-42)
written
is
A
4- ^'
-l
+ s-^-^
+ 2 + s + +~(s +
+ ^3
l)2T-(7+Tp
,
'
'
(2-43)
K =
K- 2
=$
[sX(s)] s=0
[(j
(2-44)
+ 2)-Y(j)],__ 2 =
i-
(2-45)
1) JT(j)]
|,._,=
"^
-1
<wLt(j
ls=-i
+ 2)
+ 2)2
~(2j
s 2 (y
(2-46)
-{-
2).
(2-47)
and
1
Al
d1
=2T^^ +
is +
1)
^)]
2(J
1)
^ sHs + 2)3
,
2y
d[ -2(s +
2 dsls 2 (s
(.
-i
,
"+"
2(s
l)
+ 2) 2
(2-48)
1)
F(JT2p.
24
Chap. 2
Mathematical Foundation
The completed
partial-fraction expansion
is
X = 27 + 2(7+2) ~ 7TT ~
(s
(2 " 49)
l) 3
difficult to
handle and are of special interest in control-systems studies, they deserve separate treatment here.
5= tx+jcu
Then
s= a jco
and
=
=
K...J.
K-.+j.
Example
2-6
(s
(s
+ a - jco)X(s) U_ a+ w
+ a + jco)X(s) \... . Ja
(2-50)
(2-51)
(2 - 52)
Let us assume that the values of C and are such that the nonzero poles of X(s) are
complex numbers. Then X(s) is expanded as follows
* = T + , + 1% + , f +%
(2
" 53
>
where
(2-54)
Ccu
and
oo
The
coefficients in
= coVl - C 2
(2-55)
K = sX(s)\
K-x+Ja>
(2-56)
2joo{ a +ye>)
+./
(2-57)
j(e+i/2)
2co
where
= tan-i[-f]
(2-58)
Also,
*--/,
(*
+a
+yo))A'(j)|,._
<
,. yto
(2-59)
2jco( a /a))
The complete expansion
2co
is
1
to
ejw+*n)
Sec 2 5
25
+ 5k (g-yw+*/2) e (-+y<)(
+ ~e-" sia(COt - 6)
_|_
e ne+nm e {-*-Mt\
(2-61)
or
x(t)
=
*
where 6
2.5
vi-C
~ 2 ' ~ 6)
<2
" 62
>
is
With the aid of the theorems concerning Laplace transform given in Section
2.3 and a table of transforms, linear ordinary differential equations can be
solved by the Laplace transform method. The advantages with the Laplace
transform method are that, with the aid of a transform table the steps involved
are all algebraic, and the homogeneous solution and the particular integral
solution are obtained simultaneously.
Example
Consider the
2-7
differential
x(t)
equation
?dx(t)
-d$T + 3 ~^- +
where u s (t)
The
initial
is
we first
2*(f)
is
...
= 5u,(t)
(2-63)
defined as
dx(t)/dt
|,. 0+
on both
= 2. To solve
sides
of Eq. (2-63);
we have
s*X(s)
- sx(0+) - *<(0+) +
3sX(s)
- 3*(0+) + 2*0) =
and solving
(2-65)
for X(s),
we
get
w=
*Y(s)
Equation (2-66)
is
~ sl ~ s + 5
- ~ ja ~ s + 5
+ 3s + 2) ~ s(s + 1)0 + 2)
Kt\
(2_66)
j(j*
expanded by
* *)= 2!-j4t+20T2)
(2 " 67)
Now
we
solution as
x(t)
The
first
term in the
last
5e-<
equation
is
+ le- 1
'
>
(2-68)
and the
last
two
terms are the transient solution. Unlike the classical method, which requires separate
26
Mathematical Foundation
Chap. 2
and the
If only the
theorem
may be
= lim
first
= Iim ~~f~ S ~t =
T 3S +
J-.0 ?
sX(s)
s-0
(-.00
left
of
is
Thus
applied.
lim x{t)
where we have
method
one operation.
4"
2.
(2-69)
checked and found that the function, sX(s), has poles only in the
Example
2-8
differential
equation
(2-70)
s (t)
where u s (t) is the unit step function. The initial values of x(t) and dx{t)jdt are assumed
to be zero.
Taking the Laplace transform on both sides of Eq. (2-70) and applying zero initial
conditions, we have
s 2 X(s)
we
fore,
1000X(s)
(2-71)
obtain
The
34.5sX(s)
(2 " 72)
X s) = s(s +
(
One way
17.25 -j26.5)(s
is
17.25
(2 " 73)
+ J26S)
(2-73), giving
31.6
= + 5^63) U +
1
X(s)
g/<e+s/2)
e -J(+*/2)
17.25
-/26.5
+s+
17.25
-i
+726.5J
(2 " 74)
where
tan-i(:=
J!)
-56.9
(2-75)
x(J)
Another approach
is
to
1.193e- I7
compare Eq.
CO n
2 ='
sin(26.5f
- 6)
(2-76)
= ^1000 =
31.6
(2-77)
and
C
2.6
x(t)
is
= 0.546
(2-78)
2' 6
it is
may not
Sec. 2.6
a.i*i
+
+
+
a, 2 *2
22*2
+
+
+
a 2 x 2
+ a x =y
+ a 2n X = J2
...+
ln
ax
let
us consider
=y
(2-79)
=y
(2-80)
The symbols A, x, and y are defined as matrices, which contain the coeffiand variables of the original equations as their elements. In terms of matrix
cients
which will be discussed later, Eq. (2-80) can be stated as: The product
of the matrices A andx is equal to the matrix y. The three matrices involved here
are defined to be
algebra,
21
(2-81)
a2
x2
X
(2-82)
x
>l'
(2-83)
JV
which are simply bracketed arrays of
coefficients
and
variables.
Thus, we can
Definition of a Matrix
A=
1
-2
3
In this text
we
10"
-2
A =(
\1
10
A=
-2
:)
[aj 2i3
28
Chap. 2
Mathematical Foundation
It is
Matrix
An
An
with n rows
value, although
square matrix (n
m) has a
square),
Has a
deter-
value,
minant.
definitions of matrices are given in the following.
Some important
When
Matrix elements.
au
is
As a
a matrix
is
an
a, 2
a 13
#21
fl
22
a 23
a 31
a 32
a 31
written
(2-84)
first
of
Order of a matrix. The order of a matrix refers to the total number
three
has
in
Eq.
(2-84)
matrix
the
example,
rows and columns of the matrix. For
matrix.
rows and three columns and, therefore, is called a 3 X 3 (three by three)
"
by m."
or
"n
m"
x
termed
is
columns
and
m
In general, a matrix with n rows
Square matrix.
square matrix
is
as columns.
is,
an
m X
1 matrix,
is
m>
1.
simply
Quite often, a column matrix is referred to as a column vector or
column
typical
is
a
(2-82)
in
Eq.
matrix
The
rows.
m
an m-vector if there are
matrix that is n X 1, or an n-vector.
Row
column, that
is,
row matrix
is
a 1 X n matrix.
matrix.
vector.
Diagonal matrix.
all
diagonal matrix
is
for
"
"5
0"
(2-85)
a 22
a 33 .
matrix with
Unity matrix (Identity matrix). A unity matrix is a diagonal
is often
matrix
unity
to
1.
equal
diagonal
(i
=j)
main
all the elements on the
matrix
is
unity
of
a
example
An
designated by I or U.
0"
Tl
1 =
(2-86)
1
1
Sec. 2.6
Null matrix.
all
equal to zero;
for example,
"0
o=
Symmetric matrix.
0"
(2-87)
symmetric matrix
is
satisfies
the condition
au
for
all
andj.
changed with
A
its
(2-88)
<*jt
symmetric matrix has the property that if its rows are intercolumns, the same matrix is preserved. Two examples of sym-
T6
10
the
-4".
10
(2-89)
-1
a square matrix
is
designated by
detA
As an
illustrative
A^
(2-90)
-r
-1
The determinant of A
|A|
oj
(2-91)
is
-1
|A|
3
1
= -5
(2-92)
Singular matrix.
determinant
minant,
is zero.
When
exam-
ple, let us
2x
-x,
x,
3x2 + x =
+ x + x =
- 2x + 2x =
3
Note
(2-93)
AX =
30
Chap. 2
Mathematical Foundation
where
A=
-3
-1
-2
(2-94)
X=
(2-95)
A3
and
a 3
is
null matrix.
The determinant of A
-3
|A|= -1
-2
A of Eq.
is
4-3+2-
(2-96)
rows of A are
dependent.
Transpose of a matrix. The transpose of a matrix A is defined as the matrix
that is obtained by interchanging the corresponding rows and columns in A.
matrix which is represented by
be an n X
Let
A=
Then the transpose of A, denoted by
A'
is
The transpose of A
is
is
given by
(2-98)
[a] m>
As an example of the
2-9
A',
= transpose of A =
Example
(2-97)
[%], m
n.
.0
-1
r
5.
given by
3
A'
-1
Ll
Skew-symmetric matrix.
that equals
its
negative transpose
skew-symmetric matrix
;
that
A=
Some
5.
is
a square matrix
is,
-A'
(2-99)
=A
(A)' = AA', where k is a scalar
(A + B)' = A' + B'
(AB)' = B'A'
(2-100)
1.
2.
3.
4.
(A')'
(2-101)
(2-102)
(2-103)
Sec. 2.6
where the
ting the
ij
[if
row and
Example 2-10
the jth
The adjoint
z'th
31
defined as
is
A=
n.
(2-104)
As an example of determining
a 2 x 2 matrix,
it
let
by (l) i+J
us consider
first
A = n
a2
The determinant of A
1,1 cofactor,
flu!
= an
12
a%i
o 22
#22-
is
|A|
The
fll2
is
a xl
(1, 1)
and the
element of A|,
A is
adj
1,2 cofactor
is
1,2 cofactor
2,2 cofactor.
cofactor
2,1 cofactor"
(2-105)
2,2 cofactor.
#12
#22
auj
#21
As a second example
Ie of the
a 22 the
1,1 cofactor
J ,2 cofactor
2-1
is
2,1 cofactor
1,1
Example
is
2,2 cofactor
"a ii
aj 2
an
#21
#22
#23
-O a
a 32
#3 3
(2-106)
Then
cofactor
2,1 cofactor
3,1 cofactor"
1,2 cofactor
2,2 cofactor
3,2 cofactor
cofactor
2,3 cofactor
3,3 cofactor_
1,1
adj A =
1,3
(#22#3 3
(#21#33
-
(#21#32
Chap. 2
32
Mathematical Foundation
A=
conjugate matrix of A
=
=
where a tl
2.7
(2-108)
[<y
complex conjugate of a u
Matrix Algebra
define matrix algebra in
carrying out matrix operations it is necessary to
division, and other necessary
the form of addition, subtraction, multiplication,
that matrix operations are
stage
this
out
at
point
to
operations. It is important
operations for scalar quantities.
defined independently of the algebraic
When
Equality of Matrices
Two
matrices
and
the
are said to be equal to each other if they satisfy
following conditions:
1.
2.
/and j
for every
b iJ
is,
For example,
A=
_2
implies that a n
b u a 12
,
a 12
a ti
fl 22.
b 12 a 21
,
bn
bn
Pl\
022_
(2-109)
b 2U and a 22
Addition of Matrices
Two
order.
matrices
Then
A+B=
[a l} ], m
[b,jl,,
=C=
(2-110)
[c,7 L,,
where
ClJ
for all
and/ The
Example 2-12
au
As an
illustrative
"
A=
-1
+
is
2"
B = -1
-1_
3+0
-1-1
C=A+B
_
(2-111)
b 'J
is
0^
given by
~
3
+ 3"
-2
=
4 + 2
_
-1+0.
5"
-1-
(2-112)
Sec. 2.7
.,
Matrix Algebra /
33
Matrix Subtraction
C=A-B=
_ l
= a ul,m + [-b
m
[,,]
[blJ
iJ ] n .
(2-113)
C
l tjh,m
where
Cti
for
all /
Associate
The
= a b
(2-114)
and /
still
subtraction. Therefore,
(A
B)
+C=A+
(B
C)
(2-115)
A+B+C=B+C+A
(2 " 116 >
=A+C+B
Matrix Multiplication
B=
Then
and
[*>],,,
C = AB = [a] P [b,j\q m = [c J,
The matrix C will have the same number of rows
,
(2-H7)
as
A and
as B.
may
made with
AB = A postmultiplied
AB = B premultiplied
by
by
34
Chap. 2
Mathematical Foundation
Having established the condition for matrix multiplication, let us now turn
to the rule of matrix multiplication. When the matrices A and B are conformable
to form the matrix C = AB as in Eq. (2-117), the yth element of C, c, is given by
Cii
for
1, 2,
Example
2-13
n,
andj
1, 2,
m.
A=
we notice
(2-118)
Li a.kbkj
that these
[fl,
7] 2
B=[6
; ,]
3 ,i
Thus,
"An"
Hn
U
0n
12
fln
13
-On
Oil
a 13-i
AB
bn
(2-119)
631tfllAll
a2lAu
Example 2-14
12^21
#22621
+ 013631
+ O2363I
-r
"3
A
notice that both
-3
AB
o_
and
BA
-r
-r
"i
AB =
B=
J2
we
+
+
3.
o_
o_
_2
"(3X1)
(0)(1)
_(2)(1)
_
+ (-1)(2)
+ (1X2)
+ (0)(2)
-1
-3
(3)(0)
(0)(0)
(2X0)
+ (-D(l) (3)(-l)+(-l)(0)"
(OX-D+dXO)
+ (1)(1)
(2)(-l) + (0)(0) _
+ (0X1)
(2-120)
-2_
_2
"3
-1"
_2
0_
-11
"1
BA
_2
Oj
"(1X3)
_(2)(3)
+
+
(0)(0)
(1)(0)
+
+
(-1X2)
(1)(-1)
(0)(2)
(2)(-l)
+
+
(0)(1)
(-1X0)"
(1)(1)
-1-
(0X0)
(2-121)
n 1"
1
j>
-1.
same
order.
Sec. 2.7
Matrix Algebra
36
Although the commutative law does not hold in general for matrix multiand the distributive laws are valid. For the distributive
law,
we
state that
A(B
C)
= AB + AC
(2-122)
= A(BC)
(2-123)
if
For the
associative law,
(AB)C
the product
if
Multiplication
conformable.
is
by a Scalar k
Multiplying a matrix
element of A by
by any
k. Therefore, if
A=
kA
scalar
is
[a u ] m
(2-124)
[ka u ]. m
when we
write
=y
(2-125)
x=-y
(2-126)
ax
leads to
it
or
a *y
(2-127)
Ax
then
it
may be possible
1.
2.
If
(2-128)
to write
x
where
= A _1 y
(2-129)
-1
exists are:
A is a square matrix.
A must be nortsingular.
exists, it is
given by
A-'
Example 2-15
adjA
A=
"11
-Ozi
the inverse of
(2-130)
A is
12
a 2 2-
(2-131)
given by
a 12
22
A'
adj
(2-132)
dllOll
1221
Chap. 2
36
Mathematical Foundation
where for
A to
^ 0, or
be nonsingular, A|
|
011^22
If
we pay
1221
(2-133)
is
we
2 matrix, adj
Example
is
2-16
L
the determinant of
-1
"0
A has
(2-134)
A is
IA|
Therefore,
-10
A=
A"
(2-135)
given by
is
"-2
1
-1
-2
(2-136)
lj
Some
1.
2.
3.
AA" =A"A
1
(2-137)
=A
(2-138)
(A" )"
In matrix algebra, in general,
1
AB = AC
does not necessarily imply
C.
(2-139)
easily construct
However, if A is a square
both sides of Eq.
premultiply
can
we
nonsingular,
an example
matrix, and is
Then
(2-139) by A"
1
AAB = AAC
f
IB
= IC
=C
which leads to
4. If
and
(AB)" 1
(2-140)
(2-141)
=B-'A-'
(2-142)
Rank of a Matrix
is
the
Sec. 2.7
Matrix Algebra /
"0
rank
.0
o.
"3
"0514
rank
_3
37
2,
"3
rank
2,
rank
The following properties on rank are useful in the determination of the rank
of a matrix. Given
matrix A,
annxm
1.
2.
3.
Rank of A
Rank of A
Rank of A
=
=
=
Rank of A'.
Rank of A'A.
Rank of AA'.
=2
/(x)
which
is
We
/(*)
a,jx,xj
can write
this
= 2 *,!>,,*,.
(2-143)
equation as
(2-144)
Let
= %
yt
Then Eq.
(2-144)
if
we
(2-145)
2 X&
(2-146)
becomes
/00
Now
a,jXj
define
~y\
Xi
x2
x
yi
xn _
_j_
/(*)
x'y
(2-147)
(2-145),
= Ax
(2-148)
where
A=
[atf ],.
(2-149)
38
Chap. 2
Mathematical Foundation
Finally, /(x)
becomes
/(x)
(2-150)
x'Ax
quadratic form as in
a n for i
j, given any
matrix.
In other words,
symmetric
a
with
Eq. (2-150), we can always replace A
that
such
B
matrix
symmetric
define
a
given any A, we can always
is
bn
=?ii+,
i*j
(2-151)
design algorithms.
Definiteness
Positive definite.
An
X n matrix
is
all
AI
A =
(2-152)
of A,
are positive. Equation (2-152) is called the characteristic equation
of
A.
eigenvalues
roots are referred to as the
Positive semidefinite.
its
The matrix
(n
n)
is
Negative
is
zero.
definite.
eigenvalues are
The matrix A {n x
negative and some are positive.
Indefinite.
n)
is
indefinite if
and the
some of
if all its
of a square matrix
is
to check
principal
the signs of all the leading principal minors of the matrix. The leading
matrix
square
Given
the
follows.
as
defined
are
matrix
n
minors of an n X
a 12
"an
...
0i
12
021
<*22
a lt
Then
the definiteness of
A is
A is
0n
012
013
^21
22
a23
a 31
032
"33
determined as follows:
of
positive (negative) definite if all the leading principal minors
are
positive (negative).
A is positive semidefinite if A =
|
A are
nonnegative.
and
all
_
__
z-Jransform / 39
Sec. 2.8
is
negative semidefinite
if
|
of
A|
A are nonnegative.
We may
and
all
2.8
z-Transform 1213
'*(')
%/^T)5it
kT)
(2-153)
c(kT)
Discrete-data
r^
p>
_X.
'('>
Fig. 2-3.
system
,pK
',?(')
system
Fig.
2-4.
Block diagram of a
iinite-pulsewidth sampler.
system.
A sampler is a device
data.
For example,
Fig. 2-4
'J(0
where ut)
is
'(0
[u.{t
*r)
u,(t
-kT- p)]
(2-1 54)
40
Chap. 2
Mathematical Foundation
ire)
r(t)
and output
T 2T
For small p, that is, p <^T, the narrow-width pulses of Fig. 2-5 may be
approximated by flat-topped pulses. In other words, Eq. (2- 54) can be written
1
r*(t)
(t
(2-155)
s {t
Multiplying both sides of Eq. (2-1 55) by l/p and taking the limit as p approaches
zero,
we have
Hm r*(r)
p
p-0
Urn f]
P^O
k=o
p r{kT)[u
t, r{kT)8(t
s (t
kT)
u,{f
- kT - p)]
kT)
or
lim
r*(0 - r\i)
(2-156)
we have made
6(t)
lim -L [k,(0
J>-0
u,(t
~ p)]
(2-157)
pling instant
r*(t)
r(t)
*.
V
*
Ideal sampler
Fig. 2-6.
pler
ideal
sam-
(t)
Sec. 2.8
z-Transform
41
r*(t)
T IT 3T 4T
and output
signals of
HT
an ideal sampler.
Now
R*(s)
r(kT)e-
(2-158)
the
do not have
tables
entries
that
with trans-
The z-transform
is
defined as
z
(2-159)
Equation
= In z
(2-160)
R*(s
= -L in z) =
R(Z)
is
written
r (kT)z~
(2-161)
or
R(z)
z-transform of
/*(/)
(2-162)
= [Laplace transform
of r*(t)] s=1/Tlnz
42
ap
Mathematical Foundation
Therefore,
z =
we have
'
Ts
.
2-17
= e- kT
r(kT)
where a
= 0,1,2,...
(2-163)
a constant.
Eq. (2-153),
is
From
#*(/)
= S
e~ kT5(t
(2-164)
kT)
Then
r*(s)
f; e
*=o
-' kT
e- kTl
(2-165)
-<- + )T
and subtract the resulting equation from
Multiply both sides of Eq. (2-165) by e
can be written in a closed form,
R*(s)
that
easily
show
can
;
now
we
Eq. (2-165)
*>(*)
for
<T is
itfy>
forle-^z-'l
Example
<
2-18
._.
/*(/) is
(2-168)
1.
In Example 2-17,
r(*D
if
l,
= 0, we have
* = 0,1,2,...
Biz)
This expression
(2-167)
the z-transform of
Then
" 166)
<
|g-(a+)r|
where
<2
!__}-<...
is
= z-* =
k=a
written in closed
R(z)
all
equal to unity.
= S=
A
(2-169)
Then
~
e kTs
form
(2 " 170)
(2-171)
as
y4^
i*-m<i
<
2 - 172)
or
^) = F=1
Z_1
I<1
(2
- 173
>
form
An
is
We
an
Ts
.
Sec. 2.8
z-Transform
Table 2-1
Laplace Transform
Time Function
Unit step
StU)
-e~
43
Table of z-Transforms
Unit impulse
z-Transform
S(t)
(/)
=2,6(t-
z-
nT)
z-
n=0
Tz
\_
s2
(Z
T*z{z
2
11
s n+l
1(Z
lim
n\
1)2
1)
1)3
(-1)"
d- (
n\
daAz
e-r)
U+
a) 2
a
s(s
a)
(1
e~<"
(z
(*
sin
+ CO 2
CO
a) 2
of
cos
+ CO
+a
+ a) 2 + co 2
2
z 2 e 2<,T
(s
cot
z2
e-'T)
cos coT)
2z cos co7"
z2
cot
z(z
z2
e-*r)z
l)(z
s
S2
z sin coT
2z cos coT
ze~"r sin coT
z2
co 2
e- T
te~ a <
co
S2
Tze~' T
{z
e-" T ) 2
Table 2-1 gives the z-transforms of some of the time functions commonly
used in systems analysis. A more extensive table may be found in the
litera-
ture. 12 13
-
Inverse z-Transformation
Just as in the Laplace transformation, one of the major
objectives of the
is that algebraic manipulations can be
made first in the zdomain, and then the final time response is determined by the inverse
z-transformation
z-transfor-
retained.
With
this in
2.
3.
panded by
sum of simple
is
ex-
recognizable terms,
Chap. 2
44
Mathematical Foundation
table
is
**)
For
this,
<
2 - 174)
we should
= r a + r -+-"
first
recommended procedure.
Example 2-19
_
~
R( ^
R{z
>
it is
~ e "T z T
- e)
(z
l)(z
R(z)
z
= _J
z
(2-175)
)
we have
(2-176)
e-" T
Thus,
is
found to be
r(kT)
Power-series method.
- e~'kT
(2-178)
is
the value
of z" In view of Eq. (2-161), the coefficient of z~* is
we
in
Eq.
(2-175),
R(z)
for
the
or simply r(kT). For example,
r(t) at
expand
powers
t = kT,
series in
of
(2-177)
inverse z-transform of
the z-transform table of Table 2-1, the corresponding
From
R(z)
n( .\
it
into a
21( Z )
power
(i
series in
powers of z'
-->"
_
.
(1
e-
2or
+ (l _ e - kT)z- k +
by long
)z- 2
(1
...
division; then
e~
3 T
we have
)z-i
(2
" 179
>
or
R(z)
ThUS
=S
*=
(1
'
which
e-
kT
)z-
( 2_18() )
r(kT)=l-e-
is
the
same
(2-18D
Inversion formula.
r(kT)=
^U
2ff
<f
R{z)z k -'dz
(2-182)
J J r
which
\z\
is
cT
3
a circle of radius
a contour integration along the path T, where T is
such
a value that all
of
is
and
c
z-plane,
centered at the origin in the
circle.
Sec. 2.8
z-Transform / 45
kT )
= 2jp
R(z)z k
= 2 Residues
For simple
~1
dz
Residue of R(z)z
k'1
k~1
at the pole z
at the pole z,
= (z-
=z
is
(2-183)
obtained as
z J )R(z)z"- 1 | x ,
(2-184)
Now let us consider the same function used in Example 2-19. The function
R(z) of Eq. (2-175) has two poles: z
1 and z
e~- r Using Eq. (2-183), we
have
r(kT)
[Residue of iJ(z)z*- at z
T
)z
(1
(z-e-r
=
T
(1
1]
)z
at z
= e~'T
(z-1)
,.,
[Residue of R(z)z k
(2-185)
,_,^,
= - e-" kT
1
If r x {kT)
and
r 2 {kT)
J? 2 (z), respectively,
then
gUr^kT)
2.
Multiplication
r 2 (kT)]
where a
is
(z)
as[r(kT))
2 (z)
(2-186)
by a Constant
S[ar(kT)]
3.
=R
aR(z)
(2-187)
a constant.
Real Translation
$[r(kT
nT)]
g[r(kT
nT)]
= Ar{z) -
z~"R(z)
(2-188)
and
"jj
r(kT)z- k
~]
(2-189)
where n
is a positive integer. Equation (2-188) represents the ztransform of a time sequence that is shifted to the right by nT, and
Eq. (2-189) denotes that of a time sequence shifted to the left by nT.
The reason the right-hand side of Eq. (2-189) is not z"R(z) is because
is
Chap. 2
46
Mathematical Foundation
Eq. (2-189)
Thus, the second term on the right-hand side of
to the
shifted
it
is
after
lost
is
that
simply represents the sequence
>
0.
by nT.
left
Complex
4.
Translation
Z[e*
5.
kT
R{ze^ T )
(2-190)
Theorem
Initial -Value
lim r(kT)
if
r(kT)\
(2-191)
lim R(z)
6. Final-Value
Theorem
lim r{kT)
lim
(1
(2-192)
z' l )R{z)
if
The
the function,
these theorems.
following examples illustrate the usefulness of
Example 2-20
/(/)
Let
(1
KO =
te--,
translation
0.
?>0;then
R(z)
*[/.(01
= ^jjj
g(kT)
R{ze")
(2-193)
Example
glte-'uM]
= Jl^-.ry
R( z >
(z
l)(z
0.792z 2
0.416z
S K*r> -
easily
R(z)
7 9 2Z
41 6 2
z2
i.i21z-
+ 0.981z" +
6
_o 416 z + 0.208
0.998z~ 7
R(.z) in
1.091*-'
="
'
powers of z~
1.013*"*
0.986z-=
^^
value of unity.
(2496)
+ 0.208
in
\z\ =
92
lim
checked by expanding
0.792z-i
(2 . 195)
+ 0.208)
It is
194 )
infinity.
(l-^-W)- z2 _
is
"
2-21
This result
(2
series
final
Chap. 2
Problems
47
REFERENCES
Complex
1.
wood
Cliffs, N.J.,
R. Bellman, Introduction
2.
one,
3.
C.
'
1952.
to
"dO.
Kuo,
*<?/
Ai*,*, an/
Systems, McGraw-Hill
York, 1967.
Prentice-Hall, Inc.,
Electrical Engineers
5.
New
R. Legros and A. V.
4.
Enele''
ETT
1960.
%I?
lttt
Poles,
8
'
L
tI
Partial Fraction
!f.
u7
fuby Digital
Multiple
Poles
Computer,"
pp. 161-162, Mar. 1964.
IEEE
10
'
High Order
"
IEEE
Partial
Coefficients,"
"
R R
Partkl FraCti
f:.
'
Feb. 1968.
"'
CT-11
QdCk ChCCk n
Fraction Expansion
^ix
Trans. Automatic Control, Vol. AC-11,
pp. 318-319, Apr. 1966.
UNIS
'
Coefficients of
v
?A^
yrpp.
Vol.
AC-16,
'
IEEE Trans.
p 133
'
Automatic Control,
B. C.
Kuo
Hall, Inc.,
13.
C.
Kuo,
Champaign,
Discrete
Prentice-
Illinois, 1970.
PROBLEMS
2.1.
(a)
G(s)
()
(b) G(s)
'
5(s
1)
+ 5)
+ 1)
+ 2)(s> + 3* +
*Hs
2Xs
* 2 (*
(s
2)
at infinity)
: :
48
Chap. 2
Mathematical Foundation
(C)
G(s)
(d) G(s)
f'
^u
-Ke"
=
:
(j
2.2.
DC*
+ 2)
= te-*>
= cos 5/
#(0 = e~' sin CO/
git) = S sikT)5it -
(a) g{t)
(b) git)
(c)
(d)
fcT); <*(<)
*=o
2.3.
in Fig. P2-3.
gU)
sl
(a)
irOT
l
(b)
Figure P2-3.
2.4.
fit)
t <
1<? <3
1
'
3<f<4
4
2.5.
<t
dM + 4/(0 =
wm^j
^
J
5
dt*
Assume
2.6.
dt
-,
uAt)
(a) Gis)
is
+ 2)is +
3)
(b) Gis)
(.?
(c)
Gis)
(d) Gis)
+ D 2 (s +
- s(j2
4)
10
+ 4)(j +
- s(s2 +s 2)
+
1}
Chap. 2
2.7.
Problems
49
"3
(a)
_0
"
(b)
-5_
0~
10_
- -4
3_
5.
(c)
-3
r 20"
15"
-1
10
s-3
2.8.
AB
and
rii
(a)
A=
B=
[6
1]
_3_
-r
"2
(b)
B=
0.
2.9.
set
10
-1
9"
-1
0.
+ x2 x =
*i + 3*2 x =
3*i 7* 2 2*3 =
5x,
2.10.
set
of
form (t)
+ Bu(:
*i(0
x 2 (t)
x 3 (t)
2.11.
= -*i(0 + x 2 {t)
= -2* 2 (0 - 3* (0 + in(t)
= -*,(?) - 5x2 (t) - 3* (f) +
3
(a)
5"
-1^
JO
(b)
-2
1
1
-1_
41
-l
(c)
-11
-1_
_-i
2.12.
"3
(b)
_0
"2
Ll
2"
1
3_
8"
3_
u 2 (t)
= Ax(f)
50
Chap. 2
Mathematical Foundation
2.13.
(b)
_-l
2_
"
-r
-2
_
2.14.
signals are
T seconds.
(a) /(/)
(b) fit)
2.15.
i_
The following
= te~*
= e~" sin cot
(a)
G(s)
is
+ a)n
1
(b) G(j)
s(s
5)
=
s (s + 2)
g(f) = t*e-*'
git) = sin cot
1
(c)
(d)
(e)
2.16.
CKs)
G(z)
(z
l)(z 2
expansion
+ 1)
+z
1)
3
Transfer Function and
Signal Flow
3.1
Graphs
Introduction
ponents.
by a schematic diagram
the system components. From the mathematical standpoint,
algebraic and differential or
difference equations can be used to describe
the dynamic behavior of a system
In systems theory, the block diagram is
often used to portray systems of
all types. For linear systems,
transfer functions and
among
3.2
equa51
62
/ Transfer
Chap. 3
d m r{t)
where
a au
c(r) is
dm
and
'/(?)
dr(t)
/,
r(t) is
,.
is
The
coefficients,
>
m.
b m are constants, and n
The differential equation in Eq. (3-1) represents a complete description of
the system between the input r(t) and the output c(t). Once the input and the
initial conditions of the system are specified, the output response may be
,
and b
bt
on
digital
ential equations.
A convenient way
(a s n
The
a^""
+ a ^s + a)C(s)
= (b s m + b.s'"-' +
initial
conditions,
we have
...
b m .,s
-L-
b m )R(s)
is
therefore,
-
b sm + b sm
r(*\
UW -Q~ R(s) ~ a + a.s"-
s"
+...
+...
+ bm . s + b m
+ a . s a.
1
(3 _ 3)
-,-
state
1.
2.
strictly,
of a system
3.
4.
is
is
given to illustrate
how
Example
3-1
series
RLC
designated by
is shown in Fig.
The output variable in
network
e,{t).
3-1.
this case
is
can be defined as
the voltage across any one of the three network elements, or the current
Sec. 3.2
+ o
R
v\M
i(t).
Tm
'W
ei<"
;;C
^)
written
=Ri(t)+L^ + j /(,) dt
(3-4)
e c (t)
-.
= (r+Ls+^)I(s)
EAs)
Fig. 3-1.
is
If
RLC network.
we regard
the current
function between
./(*)
E,(s)
e,{t)
(/)
and
an output
as
i(t) is
(3-4)
(3-5)
simply
Cs
53
^'^
is
E (s)=-~Ks)
c
(3.7)
E (s) _
,(*)
The
+ RCs + LCs
<3
-8
multiple
outputs.
is
individual effects.
As a
let
C,(j)
Ci(s)
= G, ,(*)*,(*) + G 12 (s)R
= G ,(s)R (s) + G 22 (s)R
2
2 (s)
(3-9)
2 (s)
(3-10)
where
2 (s)
some
reference levels.
54
Chap. 3
Since Eqs. (3-9) and (3-10) are written with the assumption that the system
is
In general,
R k (s) =
0,
l,2,...,p,
effect,
Cls)
is
tJ
It is
(s) is
(3-11)
k^j. Note
{s)
0.
defined as
is
is
defined with
= G n {s)R
R 2 (s) =
Rj(s)
+G
i2
{s)R 2 {s)
...+
z'th
output
by
G, p (s)R p (s)
(3-12)
= t Gu {s)Rj(s)
where
is,
C (s)
G,As)
with
that
j'th
if
made
(i=l,2,...,q)
where
G(s)R(s)
(3-13)
cm
C2 {s)
C(*)
(3-14)
Cq{s)j
is
a q
Rii.s)
R(*)
(3-15)
RJis)}
is
&p X
G 12 (s)
G 2i (s) G 22 (s)
'G^is)
...
G lp (s)
G 2p {s)
G(s)
(3-16)
G ql (s)
is
aq
X p
G q2 (s)
...
G(s)_
ec 3 3
'
'
3.3
55
is
G(s)
is,
(3-17)
= 8(t)
c(0
where
(3-18)
is
linear systems,
given by
g(0
-'[G(*)]
(3-19)
The
differential equation,
from minus
all practical
purposes, r(r)
is
assumed to extend
these pulses
56
Flow Graphs
Chap. 3
r(T)
r(kAr)
(b)
Fig. 3-2. (a) Input signal of a linear system, (b) Input signal represented
sum of rectangular
mated
signal.
response
is
When
is
= kAx is considered,
the system
given by
At
which
by
pulses.
r(kAx)g(t
kAx)
(3-20)
strength Ax-r(kAx).
By use of
due to r(x) is obtained by adding up the responses due to each of the impulses
from oo to +co. Therefore,
c{t)
lim
r(kAx)g(t
kAx) Ax
(3-21)
AT oo
or
c{t)
For
all
^_j(x)g(t-x)dx
for
(3-22)
< 0,
g (t
x)
is
Thus
(3-23)
applied at
t<z
0.
Or
(3-24)
Sec 3 3
the system
if r( T )
for T
<
C
T)
<*t
(3-25)
0,
W = J' *<*)*(' - t) A
(3-26)
(3-25)
is
is
and
*,
c(t)
r(t)
so
r(t) * g(t)
(3. 27 )
interpreted as
c(t)
The positions of
r(t)
and
57
written
= /'__ r(T)g(t -
c (0
Further,
now
is
g{t) in the
convolution operation
(3-28)
may be
inter-
f'
g(i>(t
T ) dx
= git) * /(/)
= git) convolves into
(3-29)
/(?)
may be caused by
simple deterioration of components due to
wear and tear, drift in operating
environments, and the like. Some systems
simply have parameters that vary
with time
a predictable or unpredictable fashion. For
instance, the transfer
characteristic of a guided missile in flight
will vary in time because of the change
of mass of the nmsile and the change of
atmospheric conditions. On the other
hand, for a simple mechanical system with mass
and friction, the latter may be
subject to unpredictable variation either
due to "aging" or surface conditions
thus the control system designed under the
assumption of known and fixed
parameters may fail to yield satisfactory response
should the system parameters
vary. In order that the system may
have the ability of self-correction or selfadjustment in accordance with varying
parameters and environment it is
necessary that the system's transfer characteristics
be identified continuously or
at appropriate intervals during the
operation of the system. One of the methods
of identification is to measure the impulse
response of the system so that design
parameters may be adjusted accordingly to attain
optimal control at all times
In the two preceding sections, definitions
of transfer function and impulse
response of a linear system have been presented.
The two functions are directly
related through the Laplace transformation,
and they represent essentially the
same information about the system. However,
it must be reiterated that
transfer
58
Chap. 3
function and impulse response are denned only for linear systems and that the
initial
3.4
Block Diagrams
Because of
its
simplicity
and
versatility,
all types.
block diagram
is
it
can be used,
how
these
Current
Controller
Turbine
torque
Control
valve
Turbine
Inlet
Mass
pressure
flow
Output
rO
Load
Hydraulic
Pump
torque
pump
Load flow
Fig. 3-3.
are
is
depicted by
its
nonlinear
Sec. 3.4
Block Diagrams
59
Kry
s(s+a)
(a)
(b)
characteristic.
input-output
system as
E (s)
t
Em (s)
E{s)
s(s
a)
(3
"3
We
shall
trol systems
for
s lg nal
ometer,
60
/ Transfer
*-
(a)
+~
Subtraction
(b)
*-
(c)
Chap. 3
e = rl
r2
+c
Addition
+~- e
=rc
(d) Multiplication
Fig. 3-5.
systems, (a) Subtraction, (b) Addition, (c) Addition and subtraction, (d)
Multiplication.
u(t)
x(t)
x(0
u(f)
x =ax + bu
x=
(a)
f(x, u)
C(s)
G(s)
(b)
Fig. 3-6.
R(s)
(c)
systems.
is
tion
x(t)
ax{t)
bu{t)
(3-31)
where
(3-32)
x(f) is
Fig. 3-6(c)
= f [x(0, u]
Sec. 3.4
Block Diagrams
61
is,
C(s)
where G(s)
G{s)R(s)
(3-33)
is
The following terminology often used in control systems is defined with reference
to the block
diagram
/(/)>
R(s)
c(t),
C(s)
b(i),
B(s)
e(t), &(s)
e(t),
=
=
reference input
= feedback signal
= actuating signal
E(s)
R(s)
C(s)
The
From
Fig. 3-7
we
write
C(s)
G(s)&(s)
(3-34)
B(s)
H(s)C(s)
(3-35)
and
The actuating
signal
written
is
6(j)
= R(s) - B(s)
(3-36)
C(s)
G(s)R(s)
G(s)B(s)
(3-37)
G(s)H(s)C(s)
(3-38)
C(s)
G(s)R(s)
R(s)
/)
r(t)
.y
ew
'
eit)
G(s)
c(t)
bit)
B(s)
H(s)
C(s)
62
system
is
Chap. 3
given by
w-
^-
(3-39)
Gis)His)
reduction process
may
transfer function of
any
linear system
We
shall
show
its
block
multivariable system
Two
block diagram of Fig. 3-8(b), the multiplicity of the inputs and outputs is
denoted by vectors. The case of Fig. 3-8(b) is preferable in practice because of its
simplicity.
MO-
-- c l
(t)
Multivariable
system
'
-*~c q (t)
(0-
(a)
Multivariable
i(0"
system
-*- c(/)
(b)
Fig. 3-8.
system.
the system
is
= G(s)(s)
&is) = R(s) - Bis)
B(s) = H(j)C(j)
C(j)
(3-40)
(3-41)
(3-42)
Sec. 3.4
Block Diagrams
S)
R(s)
80)
63
C&)
G(s)
Bfr)
Hfc)
Fig. 3-9.
Eq
(3-40)
yields
C(s)
G(s)R(s)
G(s)H(s)C(s)
(3-43)
G(j)H(j)]- G(s)R(s)
(3-44)
[I
provided that
I
G(5 )H(j) is nonsingular.
should be mentioned that although the development of the
input-output
relationship here is similar to that of the single input-output
case, in the
It
present
situation
matrix as
M(j)
Then Eq.
Example
(3-44)
is
[I
G(*)H(s)]-G(j)
(3-45)
C(j)
M(a)R(j)
(3-46)
written
3-2
G(s,=
_j_i
(3-47)
1
2J
and
H(s)
Lo
0"
i_
respectively.
The
is
is
evalu-
ated as follows:
1
G(s)H(s)
s+
2
1_
1
1
'
2J
s+j,
s
+ 2J
(3-48)
64
/ Transfer
Chap. 3
The closed-loop
transfer matrix
is
+
+
M(a)
[I
= -^
G( S )H(s)]-'GM
_1_
(3-49)
+
+
-2
2
5
2J
where
+2s+3+
A _ sS+1S|2
2 __ s 2
5s
(3-50)
5(5+1)
Thus
M(5)
s(s
1)
55
52
35 2
s{s
+95 +
1)(5
_ ~5
1_
2)
35
5(5
3.5
+
+
(3-51)
2
1).
signal flow graph may be regarded as a simplified notation for a block diagram, although it was originally introduced by S. J. Mason- as a cause-and-effect
less stringent.
is
o kJ y k
1,2,
N algebraic equations
,N
(3-52)
in the
form of cause-
and-effect relations
jth effect
N
XI ( am fr
m ^ to ./)(^
tri
cause)
(3-53)
k=l
or simply
output
This
is
the single
= 2 (gain)(input)
we must
first
When
set
of algebriac
of integrodifferential
(3-54)
form of Eq.
= S G kj (s)Yk (s)
(3-52), or
7=1,2,
(3-55)
Sec 3 5
Signal Flow Graphs /
65
12 >''-
ence of y 2 upon
An
signal
flow graphs
is
ys
The
= tf12.F1
= a 23 y
= Oi*yi
= a 2s y
+ a 32 y
+ a43 y
+ a 34 y +
+a
set
of algebraic
a 44 y 4
(3 " 58)
4 <,y 4
signal flow
<hiy 3
The
right.
first
drawn as shown in
y 3 depends upon a 23 y 2 and a 43y 4
therefore on the signal flow graph of Fig. 3-1
1(a), a branch of gain a
23 is drawn
trom node y 2 to y 3 and a branch of gain is
drawn from y t to y 3 with the
43
directions of the branches indicated by the
arrows, as shown in Fig 3-1 1(b)
;
states that
y4
66
/ Transfer
Chap. 3
(a)y 2
>4
y$
y*
>"5
=a n y + a 31 y3
l
a 43
o
^3
yi
y\
(b)>>2
=a n y\ +032^3. ^3
=fl23> 2
,
+ 43>'4
O
^5
(d)
Complete
signal
=a
24>
+ a 34>
y2
ai 2 yi
a 3 zy3- (b) yi
fl
44>"4
flow graph
= anyi + fl32>'3,
= anyi + any*,
yi
y*
= miyi
= auyi
and ends at y 4 is
of y 4 upon itself.
3.6
Summary
called a loop,
At
this point
it is
Flow Graphs
best to summarize
properties of the
1.
2.
3.
Sec. 3.7
Definitions for Signal
Flow Graphs
67
arranged from
effects
4.
5.
6.
aj
k is
3.7
Flow Graphs
is
and
reference.
p
Mg.
^/
graph shown in
satisfies
node. However,
it may be necessary to
regard nodes y 2 and/or 3 as output
y
nodes In order to meet the definition
requirement, we may simply introduce
eS
fS
,
3-12(b).
Notice that in the modified signal flow
graph
it is
in Fig.
flow graph
68
/ Transfer
Chap. 3
=y
=y
equations y 2
and y 3
of a signal flow
would
now
result.
reads
yi
which
is
different
from the
= yi +
<*\%y\
a i2 yi
(3- 59 )
a nyi
yi
Fig. 3-13.
from
Fig. 3- 12(a),
(3-60)
a 32 y 3
Vi
node.
an input node.
Since the only proper way that a signal flow graph can be drawn is from a
set of cause-and-effect equations, that is, with the causes on the right side of the
equation and the effects on the left side of the equation, we must transfer y 2
if it were to be an input. Rearranging Eq. (3-60),
the signal flow graph of Fig. 3-12 now become
originally
for
equations
the two
y%
y3
The
-y
(3-61)
"12
12
(3-62)
a 2i y 2
two equations
is
shown
y2
as an input node.
path
is
direction.
Forward path.
forward path
is
ends at an output node and along which no node is traversed more than once.
For example, in the signal flow graph of Fig. 3-1 1(d), y is the input node, and
there are four possible output nodes in y 2 y 3 j 4 and y s The forward path
t
Signal-Flow-Graph Algebra
Sec. 3.8
69
between j t and y 2 is simply the branch connected between y and y 2 There are
two forward paths between y L and y 3 one contains the branches from y t to y 2
to y 3 and the other one contains the branches from y to y 2 to y 4 (through the
branch with gain a 2i ) and then back to y 3 (through the branch with gain a 43 ).
.
The reader may determine the two forward paths between j>, and j 4
there are also two forward paths between j, and y s
Similarly,
Loop.
loop
is
is
there are four loops in the signal flow graph of Fig. 3-1 1(d). These are
shown
in Fig. 3-15.
"44
y*
Fig. 3-15.
Four loops
graph of Fig.
3-1 1(d).
yz - y - y*
3
a 12 a 23 a 34
3.8
is
y2
is
y4 y y
3
in Fig. 3-15
is
a 2i a i3 a 32
Signal-Flow-Graph Algebra
all
variable represented by a
node
is
equal to the
sum
the signals entering the node. Therefore, for the signal flow
70
Chap. 3
Node
Fig. 3-16.
as a
summing
graph of Fig. 3-16, the value of y is equal to the sum of the signals
transmitted through all the incoming branches; that is,
{
Jl
2.
all
Fig. 3-16,
"li}>2
31^3
4lj4
variable represented by a
(3-63)
Osijs
node
is
transmitted
we have
ye
yn
y%
3.
=
=
=
same
tfi6.Fi
a xl y
(3-64)
tfisJi
two
-X
y\
a-vb +
>
X.
yi
Sec. 3.9
a 12
a 23
V\
vi
"34
y3
y4
71
a
"56
<*45
45
ys
y6
a 12<*23tf34 fl 45fl56
re-
4.
3-18,
5.
whose block
may be
regarded as a simplified notation for the block diagram. Writing the
equations for the signals at the nodes &(s) and C(s), we have
diagram
is
S(j)
R(s)
C(s)
G(s)&(s)
H(s)C(s)
(3-65)
and
The closed-loop
transfer
function
(3-66)
is
equations,
C(s)
R(s)
_
1
G(s)
G(s)H(s)
^y
e(s)\_
R(s)
(3-67)
C(s)
<Xs)
-ms)
Fig. 3-19. Signal flow
For complex
formula will be introduced which allows the determination of the gain between
an input node and an output node by mere inspection.
3.9
was emphasized earlier that the construction of a signal flow graph of a physsystem depends upon first writing the equations of the system in the cause-
ical
we
shall give
two simple
illustrative
examples.
72
/ Transfer
Chap. 3
the lack of background on systems at this early stage, we are using two
networks as examples. More elaborate cases will be discussed in Chapter
where the modeling of systems is formally covered.
Owing to
electric
5,
Example
The
3-3
passive network
R, L, and
C elements
shown
in Fig. 3-20(a)
is
considered to consist of
E2 (s)
/,(,)
Ea (s)
is
= [E Js)-E (s)]Y S
= [h{s) - /,(*)]Z,(j)
= [E1 {s) - Ea (s)]Y (s)
= Z(j)/ (j)
1
(3-68)
( )
(3-69)
(3-70)
(3-71)
Y3 (s)
YAs)
|n
(s)
(a)
(*)
Z2 (s)
YAs)
Y3 (s)
work.
With the
variables
construct a signal flow graph using the loop or the node equations. For instance, in
Fig. 3-20(a), if we let Ii(s) and I3 (s) be the loop currents of the two loops, the loop
equations are
Sec. 3.9
Etn(s)
E (s)
73
(3-72)
3 (s)
(3-73)
(3 . 74)
=Z
=
h(s)
^ + zjfzjs) ^
i(s)
I Z2 (s) E
X {?)
+ z\(s) + Z4 (*) 7
'
(3-75)
(3
- 76
Now,
>
(s)
Z 2 (s)
Z2 fc)
is
not unique.
Z2 (s)
Z3 (s) + Z4 (s)
(s)
Z2 (s)
+ Z 2 (s)
(s)
Example
3-4
RLC
current in
^W =eM ~ Ri W ~ ecO
r
C dec (t) =
~1T
(3-77)
'<'>
(3-78)
(3-77)
and
(3-78)
by
and C,
respectively.
When we
have
sl(s)
i(0+)
we
(3-79)
(3 _ 80)
74
Chap. 3
WW-
nfw*
no
c
i(')Q
)
e c (t)
(a)
EJs)
EiU)
LU + R/L)
(c)
RLC network,
flow graph.
' =
E (s) =
c
7tW
y^c(0+)
(0+)
+ as +W)] lW - WTTRim EM
+ -gr/O)
(3 " 81)
(3-82)
The
signal flow
The
signal flow
alternative,
sEc (s)
we can
I(s),
as the noninput variables. These four variables are related by the equations
Sec
3.10
=5- [*/(*)]
/(*)
(3-83)
E (s) = s-^sEds)]
(3-84)
The
significance of using s~
Now,
3.10
in
Given a
signal flow
solve for
its
is
output relationship of a signal flow graph by mere inspection. The general gain
formula is
V Mj^k
M = na = *=1
A
(3-85)'
Jin
where
M = gain between y
in
and yout
yout
jln
(3-86)
of
nontouching* loops
or
A=
(sum of
all
gain products of
all
(sum of
two
possible combinations of
A*
the
is
(3-87)
may seem
fcth
which
forward path
formidable to use at
is
first
A; but
glance.
However,
in practice, systems
having a large number of nontouching loops are rare. An error that is frequently
regard to the gain formula is the condition under which it is valid.
It must be emphasized that the gain formula can be applied only between an
input node and an output node.
made with
*Two
mon
node.
76
/ Transfer
Chap. 3
Example
Consider the signal flow graph of Fig. 3-19. We wish to find the
transfer function C(s)/R(s) by use of the gain formula, Eq. (3-85).
3-5
There is only one forward path between R(s) and C(s), and the forwardpath gain is
M, =
2.
There
is
Pu =
3.
is
-G(s)H(s)
-/,,=
By use of Eq.
G(j)#(j).
C(s)
R(s)
result
= Mi At =
A
system
is
obtained as
G(s)
1
(390)
G(s)H(s)
Consider, in Fig. 3-20(b) that the functional relation between Eia and
E is to be determined by use of the general gain formula. The signal
flow graph is redrawn in Fig. 3-23(a). The following conclusions
3-6
1.
(3-89)
There are no nontouching loops since there is only one loop. Furthermore, the forward path is in touch with the only loop. Thus Ai = 1, and
A=
Example
(3-88)
G(s)
There
Eia
is
3-23(b).
The forward-path
gain
and
as
shown
M, = FiZ2 y3 Z4
2.
in Fig.
is
(3-91)
loop gains
3-23(c); the
are
P n = -Z2 y,
21 = -Z2 r3
i> 31 = -Z4 r
(3-92)
/>
(3-93)
(3-94)
3.
There
is
in Fig. 3-23(d);
the loop
-Z Y
2
-Z4 Y
and
Thus
Pu =
4.
(3-95)
}
0,
Eq. (3-86),
A=
=
+P +P )+P
-(J
+ Z Y + Z Y + Z4 Y + Z Z4 Yt Y
2
3l
zl
l2
2
(3-96)
Sec. 3.10
(a)
Oh
O
(b)
(d)
Ein
nontouching loops.
5.
All the three feedback loops are in touch with the forward path ; thus
A,
(3-97)
we
Substituting the quantities in Eqs. (3-91) through (3-97) into Eq. (3-85),
y,
Example
3-7
r3 z2 z4
Y
+ Z<Y + Z1 Zt Y Y
2
Ai
+Z
1 Y
+Z
is
desired for
Ec
It is
lt
obtain
(3-98)
is
linear, the
The gain between one input and one output is detergain formula to the two variables while setting the rest of the
The
signal flow
and
graph
(d), respectively.
78
/ Transfer
Chap. 3
(0+)
ec
(0+)
<(0 +)
o
"
o-
-O
/
sl
(c)
graph of the
RLC network
Forward path between Ei and /. (c) Forward path between /(0+) and
(d) Forward path between e c (0+) and /.
/.
Sec. 3.10
The
signal flow
A=
79
A is given by
1
+#*-'
L + LC
(3-99)
'
= 0,
/(0+)
-,
0,
'(0+)
<HL)sec (0+)
When
all
for
all
cases
we have
= 0,
i(0+)
we
* c (0+)
(3-100)
e c (0+)
(3-101)
E =
(3-102)
write
(3-103)
output variable,
Lc
c is
considered as the
we have
is
(3-104)
Example
3-8
Consider the signal flow graph of Fig. 3-25. The following inputoutput relations are obtained by use of the general gain formula:
yi
>-i
y_3
yi
+ dy
A
_ ag(l + d) +
A
a(i
(3-105)
abc
(3-106)
where
A=
+eg + d +
bcg
+ deg
(3-107)
3-8.
80
/ Transfer
3.11
Chap. 3
Because of the similarity between the block diagram and the signal flow graph,
the general gain formula in Eq. (3-85) can be used to determine the input-output
relationships of either. In general, given a block diagram of a linear system we
can apply the gain formula directly to it. However, in order to be able to identify
all the loops and nontouching parts clearly, sometimes it may be helpful if an
equivalent signal flow graph is drawn for a block diagram before applying the
gain formula.
To
illustrate
how
the signal flow graph and the block diagram are related,
since a
is
shown
Note that
summing point of all
in Fig. 3-26.
interpreted as a
(a)
(b)
flow graph.
(b)
Equivalent signal
Sec. 3.12
G G2 G
C(s)
R(s)
-j-
Gj
+ G G H + G G H + G G G + G H2 +
l
(3-108)
G,G 4
Similarly,
E(s)
R(s)
3 (s)
__..
G,(? 2 ff,
+ G G H + G4 H
2
(3-109)
A
_
~
+ G G H + G,H
2
(3-110)
R(s)
where
A=
3.12
+ G^H, + G G H +
2
G,G 2 C 3
+ G4 H + G Gi
2
(3-111)
It is
shown
in Fig. 2-5. Since for a very small pulse duration p, as compared with the
sampling period T, the finite-pulsewidth sampler can be approximated by an
ideal sampler connected in cascade with a constant attenuation p, the system
of Fig. 3-27(a) may be approximated by the system shown in Fig. 3-27(b).
r{t)
R(s)
-?
c(t)
*0)
G(s)
R?(s)
(P)
C(s)
(a)
KO
R(s)
-*
r*(t)
* P*
G(s)
(s)
Ideal sampler
(b)
an
c(0
>-*<,t)
'
C(s)
82
/ Transfer Function
Chap. 3
c*(t)
,y\
|"
>
rit)
c(t)
/*(/)
<*
G(s)
'
R(s)
_^_
C*(s)
R*(s)
C{s)
an
ideal sampler.
it is
is
included in the transfer function of the process, G(s). Therefore, the block
diagram of Fig. 3-28 is considered as that of a typical open-loop discrete-data
or sampled-data system.
assume that
r*(t),
the output
of the ideal sampler 5",, is a unit impulse function. This may be obtained by
or if r(t) is a unit impulse funcsampling a unit step function u,(t) once at t =
tion.* Unless stated otherwise, the samplers considered in the remaining portion
of this text are ideal samplers.
The output of G{s) is the impulse
g{i). If
fictitious ideal
sampler
S2 which
,
synchronized with Sj and has the same sampling period as that of 5,, is placed
at the output of the system as shown in Fig. 3-28, the output of the sampler S z
is
may
be written as
c*(t)
g*(t)
t g(kT)S(t ~ kT)
(3-112)
A=
where c(kT)
= g(kT)
is
function
is
G*(s)
[g*(f)]
= S
(3-113)
g(kT)e-*T-
is
is
is
When
is
weighting sequence impulse train gives the pulse transfer function G*(s).
Sec. 3.12
Once
at
r(t),
c(t),
Consider that an arbitrary input r(t) is applied to the system of Fig. 3-28
0. The output of the ideal sampler is the impulse train,
=
A=
r*(f)
By means of superposition,
c(t)
At
the system,
input
is
83
r(0)g(t)
r(kT)S(t
kT)
(3-114)
+ r(T)g(t -
T)
r(kT)g(t
is
due to
kT)
l)T]g(T)
r*(t), is
(3-115)
c(kT)
r(0)g(kT)
r{T)g[{k
1)7]
r[(k
r(kT)g(0)
(3-116)
where
it is
assumed that
system so that
its
<
0, since
the process
is
a physical
Multiplying both sides of Eq. (3-116) by e~ kTs and taking the summation
,
from k
to
c(kT)e- kT
oo,
r(0)g(kT)e- kT
we have
<t
r[{k
c(kT)e- kT <
r(T)g[(k
l)T]g(T)e-^
*=
is
l)T]e~ kT
...
+ *=0
r(kT)g(0)e^
r (T)e~ T
r{2T)e~>
is
(3-117)
simplified to
(3-118)
or
c(kT)e- Ts
r{kT)e- kT *
g(kT) e
- kT >
(3-119)
Therefore, using the definition of the pulse transfer function, the last equation
is written
C*(s)
which
is
shown
in Fig. 3-28.
R*(s)G*(s)
(3-120)
The z-transform
c(kT)z- k
relationship
Ts
,
is
Eq. (3-119)
r{kT)z' k
is
also written
g{kT)z~ k
(3-121)
G(z)
= g{kT)z' k
k=0
(3-122)
which implies that the z-transfer function of a linear system, C(z), is the ztransform of the weighting sequence, gikT), of the system. Equation (3-121)
is
written
C(z)
R{z)G(z)
(3-123)
84
/ Transfer
Chap. 3
discrete-data system is
important to point out that the output of the
of the output,
transform
pulse
the
However,
continuous with respect to time.
of c{t) only
values
the
specify
C(z),
output,
the
C*(s) and the z-transform of
well-behaved function between sampling
at the sampling instants. If c(t) is a
description of the true output:c{t).
instants c*(0 or C(z) may give an accurate
sampling instants, the zHowever if c(t) has wide fluctuations between the
only at the sampling instants, will
transform method, which gives information
It is
C*(*)
where
=i S
C(s+jnco
O 124
s)
2njT.
second and ca s
the sampling frequency in radians per
c(t)
output
continuous-data
Fig. 3-28, the Laplace transform of the
co s is
From
C(s )
is
(3-125)
G(s)R*(s)
C*(s)
We
+ jnaJR^s + jnw,)
G(s
can write
+ jnco ) = k=0
J r(kT)e-"^
R*(s
and
(3-127)
-jnkTo>,
-j2xnk
(3-128)
becomes
+ jnco =
R*(s
Using
n,
Thus Eq.
(3-127)
r(kT)e- kTs
E
fc
A:
=
since for integral
(3-126)
this identity,
Eq. (3-126)
s)
is
(3
J*(5)
" 129
simplified to
C*(s)
= i{*(4 S
C*(j)
= J?*WG*(*)
G(5+7t
(3-130)
s)
3431)
where
G*(j)
by use of z
is
is
4 S
J
in z of Eq. (3-123)
e
(3-132)
G^+jnco,)
= oo
Ts
.
.
In conclusion,
we
note that
when
given by
C(s)
If the continuous-data
output
is
(3-133)
G(s)R*(s)
is
synchronized with
Sec. 3.12
85
and has the same sampling period as the input, the Laplace transform of the
discrete-data output is given by
C*(s)
The
is
G*(s)R*(s)
natural, since
it is
(3-134)
The expression
in Eq. (3-134)
can be interpreted as
being obtained directly from Eq. (3-133) by taking the pulse transform on both
sides of the equation. In other words, in view of Eq. (3-129), we can write, from
Eq. (3-133),
is
C*(s)
=
=
it
[Gis)R*(s)]*
(3-135)
G*(s)[R*(s)]*
R*(s)
(3-136)
The
elements
system which contains two cascaded elements. In the system of Fig. 3-29(a), the
is
"2
(0
T
D*
R*(s)
d(t)
d*(t)
(s)
D(s)
D*(s)
c(t)
G 2 (s)
C(s)
w)
(0
C*Cs)
Kt)
R(s)
**
r*(t)
dU)
Gi(s)
R*(s)
D(s)
c(t)
G 2 (s)
C(s)
(b)
separates the
two elements,
and sampler
86
Chap. 3
same period as the sampler Si. The two elements with transfer functions
and G 2 (s) of the system in Fig. 3-29(b) are connected directly together. In
discrete-data systems, it is important to distinguish these two cases when deriving
the
(s)
Let us consider
first
D(s)
C(s)
=G
Gi(s)R*(s)
is
written
(3-137)
is
(s)D*(s)
(3-138)
C(s)
=G
(s)GKs)R*(s)
C*(s)
(3-139)
last
Gf(s)G*(s)R*(s)
(3-140)
in Eq. (3-136).
is
=G
C(z)
equation gives
The corresponding
(3-141)
2 (z)Gi(z)R(z)
equal to the product of the z-transforms of the two individual transfer func-
tions.
C(s)
last
Gi{s)G 2 (s)R*(s)
equation
C*(s)
in Fig. 3-29(b) is
(3-142)
is
[Gi(s)G 2 (s)]*R*(s)
(3-143)
where
[Gi(s)G 2 (s)]*
= -i
f)
G (.s+jmo.yG 1 (.s+jna>.)
1
(3-144)
[Gi(s)G 2 (s)]*
= GiG^s)
= G GHs)
(3-145)
Then Eq.
(3-143)
becomes
C*(s)
GiG^(s)R*(s)
(3-146)
C(z)
GiG 2 (z)R(z)
(3-147)
G 2 (s),
Sec. 3.12
It is
87
G Gt(s)^Gf(s)GKs)
(3-148)
and
G.G^z)
Therefore,
we conclude
sampler in between
is
^ G^Gziz)
(3-149)
is
C(s)
G(s)E*(s)
(3-150)
"X,
t)
R(s)
^
_\y
e(t)
E(S )
V
j.
c*(t)
C*(s)
c(t)
e*(t)
G(s)
E*(s)
C(s)
H(s)
R(s)
error function
is
H(s)C(s)
(3-151)
E(s)
R(s)
G(s)H(s)E*(s)
(3-152)
E*(s) gives
E * (S) _
The output transform C(s)
we have
is
R*(s)
(3-153)
G(s)
C(s)
1
,R*(s)
GH*(s)
on both
(3-154)
G*(s)
^-TTmwr
{s}
(3-155)
88
/ Transfer
In this case
it is
Chap. 3
The
G*(s)
shall
show
(3-156)
GH*(s)
is
_ G{z)
~ + GH(z)
C(z)
R(z)
We
(3 " 157)
although
it
is
possible to define a
transfer function for the closed-loop system of Fig. 3-30, in general, this
may
not be possible for all discrete-data systems. Let us consider the system shown
in Fig. 3-31. The output transforms, C(s) and C{z), are derived as follows:
C(s)
E(s)
=
=
G(s)E(s)
R(s)
(3-158)
H(s)C*(s)
(3-159)
~>"
C*(t)
KO
R(s)
(%
~?)
e(t)
\J
E(s)
c(t)
G(s)
C(s)
c*(t)
Hds)
C*(s)
C{s)
G(s)R(s)
G{s)H(s)C*{s)
sides of
the last
GH *{s)
+ **L
(3-160)
(3-161)
Note that the input and the transfer function G(s) are now combined as one
function, GR*(s), and the two cannot be separated. In this case we cannot define
a transfer function in the form of C*(s)/R*(s).
The z-transform of the output is determined directly from Eq. (3-161) to be
= r GR(z)
GH(z)
C(z)
where
it is
(3-162)
GR(z)
#[G(*)fl(s)]
(3-163)
GH(z)
g[G(s)H(s)]
(3-164)
and
Problems
Chap. 3
To determine
we
We
89
substitute
have
- G+gh*\s) GR * (s)
G(s)R(s)
(3 " 165)
and
without
difficulty, for
may become
tedious.
may
7
refer to the literature.
REFERENCES
Block Diagram and Signal Flow Graphs
1.
T. D. Graybeal, "Block
S. J.
S. J.
Properties of Signal
Flow Graphs,"
Sept. 1953.
Signal
Flow Graphs,"
L. P. A.
B. C.
New
York, 1967.
6.
B. C.
Prentice-
Hall, Inc.,
8.
B. C.
Kuo,
Champaign,
Discrete
Illinois, 1970.
PROBLEMS
3.1.
The following
where
(b)
3.2.
r(t)
c{t)
*%P +
104g->
2cit)
r(
2)
is
shown
in Fig.
90
Chap. 3
G0)
0"
.0
1.
HO)
Find the closed-loop transfer function matrix for the system.
/\
Rfe)
C(s)
,
G(s)
H(s)
Figure P3-2.
3.3.
is
shown
in Fig. P3-3.
C2 (s)
C2 (s)
Ci(j)
Ri(s)
*i(*)
G(s)R(s)
R 2 (s)
Figure P3-3.
3.4.
Draw
set
+ x 2 + 5x =
+ 2x 2 4x = 2
x 2 x =
3
of algebraic equations:
Problems
Chap. 3
3.5.
Draw an equivalent
signal flow
in Fig. P3-5.
91
Find
Figure P3-5.
3.6.
Find the
gains,
y 6 /yi,
y%ly\,
signal flow
P3-6.
Figure P3-6.
3.7.
Find the gains y<,\y x and y 2 lyi for the signal flow graph shown
-0.5
Figure P3-7.
in Fig. P3-7.
92
/ Transfer
3.8.
Chap. 3
In the circuit of Fig. P3-8, e s (t), e^t), and i,(t) are ideal sources. Find the value
of a so that the voltage e (t) is not affected by the source ed (t).
o+
Figure P3-8.
3.9.
in Fig. P3-9(a)
and
Oyj
(a)
(b)
Figure P3-9.
3.10.
Given the signal flow graph of Fig. P3-10(a) and the transfer functions G it G 2
G 3 G 4 and G 5 find the transfer functions GA GB and (7C so that the three
systems shown in Fig. P3-10 are all equivalent.
,
(a)
Figure P3-10.
Chap. 3
Problems
93
(b)
(c)
Construct an equivalent signal flow graph for the block diagram of Fig. P3-11.
(a) Evaluate the transfer function C/R when
0. (b) Determine the relation
among the transfer functions G u G 2 , G 3 G4
u and 2 so that the output C
is not affected by the disturbance signal N.
N=
Figure P3-11.
3.12.
multivariate system
is
relations
C(s)
S(.y)
= G(s)S(s)
= R(.r) - H(*)CO)
where
C(s)
= c^sy
R(s)
-R^sT
-Ri(s).
G(s)
n
1
H( S)
_1_
'I
0"
.0
0_
94
/ Transfer
Chap. 3
(a)
by using
M(s)
(b)
3.13.
[I
M(.$)RO)
G(j)H(5)]-'G(i)
Draw a signal flow graph for the system and find M(s> from the signal flow
graph using Mason's gain formula.
Find the transfer function relations C(s)/R(s) and C(s)jE(s) for the system
shown
in Fig. P3-13.
Figure P3-13.
3.14.
in Fig.
P3-14.
r(t)
c(t)
r*(t)
s(s
+ 2)
T
Figure P3-14.
3.15.
KO
cd)
r*(.t)
(a)
r(t)
^
T
r*(t)
^
s+
-X1
(b)
Figure P3-15.
c(t)
s +
in
4
State-Variable Characterization
of
4.1
Dynamic Systems
system
relations.
For
when one is intime-domain solution, which depends to a great deal on the past
history of the system, the transfer function does not carry all the
necessary
terested in a
js
An
is
is
in its
to linear systems
The
method of describing a
state-variable representation
It
linear system
not limited
can be applied to nonlinear as
is
The
state-variable
method
is
often referred to as a
modern approach.
However, in reality, the state equations are simply first-order differential equations, which have been used for the characterization of dynamic
systems for
many years by physicists and mathematicians.
To
As the word
96
State-Variable Characterization of
Cna P- 4
Dynamic Systems
the past, present, and future conditions of the system. It is interesting to note
that an easily understood example is the "State of the Union" speech given by
the President of the United States every year. In this case, the entire system
encompasses all elements of the government, society, economy, and so on. In
variable
by a
set
state equation.
set
of variables,
characteristics of a
., x(t) is chosen to describe the dynamic
Xl (t), x 2 (t),
of
the system. Then
variables
state
the
system. Let us define these variables as
conditions
following
the
these state variables must satisfy
.
1.
At any time
define the
2.
x 2 (t a ),
),
initial states
x(t
initial time.
>
Therefore,
we may
a minimal
set
output variable
is
Example
R
-^AM
4-1
in Fig. 4-1.
e(f)
i(0
fJ
RL
network for
Fig. 4-1.
state vari-
RL
network shown
The history of the network is
>
network.
E(S ) =*h.
is
^=R
(t)
+ L dm
we
get
(4-2)
m~
.,
The
current
/(/)
for
is
E,
s{R+Ls)'r
1-/(0+)
R+Ls
(4-3)
Sec. 4.2
97
We have
'(0
Once
Dynamic Equations
is
= ^ (1 - e-1 +
determined for
same time
i(0+)e- R,/L
-)
> 0,
(4-4)
is
it is
to see that the voltage across a capacitor also qualifies as a state variable.
4.2
The
equation of Eq.
first-order differential
between the
(4-1),
e{t),
=fix
=
.
(f),
x 2 (t), ...,
r p (t)]
(4-6)
1,2, ... ,n
.
xn (t)
z'th
/-,(/),
r 2 (t),
rp {t) are
functional relationship.
variables
c k {t)
&[*,(*),
x 2 (t),
...,
x(t),
rM
r 2 {i),
..., r p (t)]
(4-7)
k=\,2,...,q
where gk denotes the kth functional relationship. The state equations and the
output equations together form the set of equations which are often called the
dynamic equations of the system.
Notice that for the state equations, the left side of the equation should
first derivatives of the state variables, while the right side should
have only the state variables and the inputs.
Example
4-2
Consider the
tional
RLC
network shown in Fig. 4-2. Using the convennetwork approach, the loop equation of the network is
written
e{t)
Ri(t)
We notice that
+L^ + [
RLC network.
dt
(4-8)
this
equation
Fig. 4-2.
i{t)
is
98
State-Variable Characterization of
Dynamic Systems
Chap. 4
*i(0
= iff)
Xl(t)
(4-9)
= Rx (t)+L
eit)--
(4-10)
f /(/) dt
we have
+ ^x 2 (t)
(4-11)
Rearranging the terms in Eq. (4-11) and taking the time derivative on both sides
we have the two state equations of the network,
of Eq. (4-10),
= -T*'<'>-nr*<'> + r*>
^- =
}
which are
(4 " 12)
(4-13)
*,(/)
We have demonstrated how the state equations of the RLC network may
be written from the loop equations by defining the state variables in a specific
way. The objective, of course, is to replace Eq. (4-8) by two first-order differential
equations. An alternative approach is to start with the network and define the
state variables according to the elements of the network. As stated in Section
4.1, we may assign the current through an inductor and the voltage across a
capacitor as state variables. Therefore, with reference to Fig. 4-2, the state
variables are defined as
*i(o
(4-i4)
*2
efy)
(4-15)
=
(0 =
Then, knowing that the state equations would have to contain the first derivatives of x^t) and x 2 (t) on the left side of the equations, we can write the equations directly
by inspection from
Fig. 4-2
^p =
Voltage across L:
Current in C:
C^2 =
-Ri(t)
e e (t)
<?(0
(4
"
16 )
(4-17)
i(t)
at
**M = - Xl
4*M =
^ Xi
(f)
- i-*
(0
two equations
e(t)
(4-18)
(4-19)
(f)
which are the state equations of the RLC network. We notice that using the
two independent methods, the only difference in the results is in the definition
of the second state variable x 2 (t). Equation (4-19) differs from Eq. (4-13) by
the factor of the capacitance C.
From
these
Sec. 4.3
dxt)
% a u xj(0 +
dt
kJ Xj(t)
j=
coefficients.
+ m^
t,
i=l,2,....,n
2l b ik r k (i)
are written
(4-21)
(4-20)
k=l,2,...,q
e km r m {i)
99
coefficients of Eqs.
"*i(0'
x 2 (t)
x(r)
where
x(t)
is
1)
(4-22)
(px
1)
(4-23)
(qX
1)
(4-24)
(n
>i(0"
r 2 (t)
r(/)
WO.
where
r(/) is
"c,(0"
cz(t)
c(0
_c,(0_
where
c(r) is
Then
can be written
M)=f[x(0,r(0]
where
denotes an n
f as elements,
c(r)
where g denotes a q x
gq
as elements.
(4-25)
g[x(r), r(r)]
column matrix
(4-26)
g u g 2 ,...,
100
State-Variable Characterization of
For a
Chap. 4
Dynamic Systems
State equation
Output equation:
where
A is an
^
c(/)
+ Br(r)
(4-27)
= Dx(0 + Er(/)
(4-28)
Ax(0
a 11
fl
d 21
#22
a,
"In
12
a 2n
A=
B is
an n X
(4-29)
a,
t>u
b 12
bn
b2 2
p matrix given by
a.;
(4-30)
Pi
D is
a q
b 2
n matrix given by
~d u
d i2
du
"21
"22
d2
D=
and
is
a q
(4.31)
dql
dq2
i2
en
e 12
p matrix,
'e
"2p
E =
(4-32)
Zql
Example
4-3
The
C2
and
dxi(ty
dt
dx 2 (t)
r
'
R
L
V
'
j_
-\
r*iw"
+
\x
dt
L
e(0
(4-33)
{t)
2
C
Thus the coefficient matrices A and B are identified to
_R
A=
be
L
(4-34)
J_
Sec. 4.4
B= L
4.4
(4-35)
is
^
Let
<(f)
must
x n matrix
be an n
Furthermore,
let
(4-36)
equation
satisfy the
= Ax
homo-
= AKO
it
(4-37)
initial state at t
= 0; then
<f>(t) is
also defined
x(t)
which
is
<Kr)x(0+)
homogeneous
fy(t) is
(4-38)
>
0.
sX(s)
x(0+)
we
= AX(s)
(4-39)
get
- A)" >x(0+)
where it is assumed that the matrix (si A) is nonsingular.
X(s)
(si
x(0
Comparing Eq.
to
,- [(si
last
(4-40)
equation yields
A)" ']x(0+)
>
(4-41)
is
identified
be
(>(0
An
alternative
way of
- [(jI-A)-']
1
(4-42)
solving the
x(0
for
> 0, where e
At
It is
easy to
represents a
AI
show
=1+
At
power
4-
'x(0+)
series
^A*/ 2
is
(4-43)
+1A
a solution of the
(4-44)*
homogeneous
state
^=
*It
this
power
series is
Ae Kt
uniformly convergent.
(4-45)
102
Cha P- 4
Dynamic Systems
State-Variable Characterization of
expression for
eK<
=I+
+ ~A
At
+ 1A
2 2
/
directly
from Eq.
(4-46)
(4-42).
This
as
is left
an
Example 4-4
Consider the
that
is,
e(t)
RL
network of
is
written
*=-*)
The solution of the last equation for t
Thus
>
this case is
0(0
which
is
i(t)
(4-47)
a scalar and
e-*" L
(4-48)
is
given by
>
= 0.
(4-49)
is
Since the state transition matrix satisfies the homogeneous state equation,
represents the free response of the system. In other words,
it
response that
and
As
is
excited
by the
the
initial
matrix
it
governs the
name
The
state transition
matrix
1.
<J>(0
<f)(0)
Proof:
f
(4-50)
0.
2.
0-(O
Proof:
(4-51)
4(-O
<K0e"
At
e x 'e- A '
e- A '
by
we
get
(4-52)
_1
<J>
Ac
(')>
we 8 et
(f,-i( f )
(4-53)
e~ At
(4-54)
is
Thus
<K-r)
An
interesting result
from
<f~'(0
this property
of
<f>(t)
arranged to read
x(0+)
<J>(-0x(0
(4-55)
Sec. 4.5
State Transition Equation /
103
time.
3.
<J>(7 2
0<K>i
<f>(t 2
t )
for any
tQ
(4-56)
t2
Proof:
<K'2
'i)<K'l
= A
e
A(
"-')
=e
= M2 ?
<'-">
A(r,-)
(4.57)
t )
= <K>2 - tMt
*('.) = ftr, )x(t
x(t
= <K?2 - )x(t
*(h)
tQ
2)
The proper
result is obtained
comparing the
4
result
(4-58)
(4.59)
(4 _ 60)
by substituting Eq.
(4-59) into
Eq
(4-58)
and
MO]" =
<Kkt)
for
= integer
(4-61)
Proof:
oAr
[<t>(t)Y
(k terms)
e kAt
= Mt)
4.5
(4-62)
The
geneous
input voltage
RL
is
>
0.
104
State-Variable Characterization of
Dynamic Systems
Chap. 4
&j& = Ax(0 +
can be solved by using either the
Br(r)
(4-63)
classical
sX(s)
x(0+)
AX(s)
BR(s)
(4-64)
vector evaluated at
initial state
we have
0+. Solving
for
X(s)
The
(si
state transition
A)" 'x(0+)
(si
A)" 'BR(s)
(4-65)
is
"'[(5l
The
time
is
+ Jf
<K0x(0+)
is
<f>(?
by
state
We
x(/
),
start
initial
(4-66)
is
>
useful only
(4-67)
when
the initial
is
it
time be represented by
<J>(-/ )x(/o)
<(>(f)
0.
x(0+)
A)- BR(s)]
T)Br(r) dx
more
written
state transition
defined to be at
JC-'Kil
A)" ]x(0+)
ta
applied for
r(f) is
<K-'o)
flexible initial
time must
>
initial
r.
f"
Wo -
T)Br(t) di
get
(4-68)
x(0
(J>(0<K->o)x('o)
<KO<K-'o) Jf<Wo
T)Br(T)rfr
,,
(4-69)
<(>(?- T)Br(r) di
J
Now
last
two
integrals,
<f>(t
)x(t
<j)(t
~ r)Br(r) dx
Once
from Eq.
c(0
Dcf>('
is
(4-70)
0.
initial state
and the input vector simply by subThus the output vector is written
)x(t
f ' D<f>('
T)Br(r) dx
'
Er(f)
(4-71)
Sec. 4.5
State Transition Equation
105
equation.
Example
4-5
dt
dx 2 (t)
-2
dt
The problem
t
> 0; that
is,
is
(4-72)
>
x(0 for t
when the input r(/)
u s (t). The coefficient matrices A and B are identified to
be
r(t)
#(/)
1
|*2(0
B=
-3.
for
~0~
r
_
(4-73)
_1_
Therefore,
0~
"
_-2
1"
The matrix
inverse of (si
A)
3s
state transition matrix of
-3.
A is found
3_
1"
(4-74)
-2
-2e~'
The
#(/)
>
<
2fi-('-')
-2g-(r-r)
-2c"'
+
c
<?-'
alternative, the
'
2t
(4-75)
B and
'
x(0+)
2e~ 2 \
-(i-t)
_ e -(f-T)
c -2 '
2e~ 2
-
'
'
2 e -2('-r)
2e~'
x(0
e' 2
e~' + 2e~
e~ !
'
e~ 2 '
e -2(t-t)
_|_
e' 1
+ 2e~ 2
is
We have
2e~' ~ e' 2
e~'
2e-' + e~ 2
e -< +
'
As an
Thus
x(0
_2
"1
is
(si- A)-'
The
-1
~s
si
g-'
<
-e-<
-2<-t)
_|_2e- 2
~2'
('-'>
rfT
(4-76)
'
x(0+)
2e" 2 '.
(4-77)
>>0
e -2t
~ Ksl1
A)-iBR(s)]
=-i
's
-
= -'
-2
1"
s.
-J_-l
1
3s
(4-78)
1_
- e~' + \e~
e~< e -2
'
2>
/>0
106
State-Variable Characterization of
Example
Dynamic Systems
In this example
4-6
method
Chap. 4
we
RL
is
as
shown
in Fig. 4-4.
IE
e(t)
The
state
waveform
for the
network
in Fig. 4-1.
is
di(t)
rf-
= -r /(/ + rw
(4-79)
Thus
A
The
R
L
to the
(4-80)
is
#(0
One approach
B
e~ R " L
(4-81)
;'(?)
for
?,)
>
is
voltage as
e(t)
where u s (t)
is
Eu s (t)
Eu
(t
(4-82)
is
(s)=^ (!+*-")
(4-83)
Then
(si
- A)-'BR(j)
Rs[l
R/L L
E
+ (L/R)s] (1
*
e- R "H(0+)u s (0
the input
to
state transition
t
t,,
and
d+e-"')
+ d -
>
(4-84)
+<?-'")
is
obtained:
e-"*)u,(t)
(4-85)
e -M-w-]u.(t-t 1 )
?,
to
Eu,(t)
<
<,
t,,
is
e(t)
0<t <
r,
(4-86)
Then
(4-87)
j
Rs[l
+ (L/R)s]
Relationship
Sec. 4.6
i(0
Substituting
i( tl )
The value of
period of
< <
t
we
i(t) at t
e -Ru/L i( Q
is
ti is
now
(4-88)
+ ) + -j| (1 -
used as the
e -*"> L )
(4-89)
initial state
< < co. The magnitude of the input for this interval
t
tt
i(0
107
get
where
tx
is
2E. Therefore,
is
>
(4-90)
t,
i(ti) is
This example illustrates two possible ways of solving a state transition problem.
In the
first
Although the
is
is
treated as
in
more
method
first
4.6
In preceding sections
we
it
is
always possible to
it is
necessary to investigate
how
state equations
can be written directly from the differential equation or the transfer function.
The relationship between a high-order differential equation and the state equations
is
is
de-
where
dn
c(t) is
~l
c(i)
d"~ 2 c(t)
The problem
is
dc(i)
is
,..
,,..
lA ,,.
the input.
by n
state equations
and an output
equation. This simply involves the defining of the n state variables in terms of
the output c(0
and
its
derivatives.
We
have shown
convenient
is
state variables as
met.
we
108
Dynamic Systems
State-Variable Characterization of
it is
Chap. 4
x,(0
c(t)
*a(0
dt
(4-92)
*(o
fife""
Then
dx z (t)
dt
=*
(0
(4-93)
<fa-i(0
xJLO
<//
dx(t)
= -a x,(0 B
dt
a_,x 2 (0
...
a 2 Jf-i(0
i^(0
+ KO
where the last state equation is obtained by equating the highest-ordered derivative term to the rest of Eq. (4-91). The output equation is simply
c{t)
where
x(t) is the
x,(0
is
written
= Ax(0 +
.
state vector
and
(4-94)
Br(t)
r(i) is
(4-95)
The
coefficient
matrices are
1
n)
(4-96)
a-i a
n -2
A,-3 fl_4
0"
(xl)
(4-97)
Sec 4 7
in vector-matrix
c(0
form
is
= Dx(r)
(4-98)
where
D=
The
state
...
[1
0]
(1
is
and B defined as in
called the phase-variable canonical form in the next
(4-99)
ji)
section.
Example
4-7
last
The
(4100)
is
equated to
*i(')
= c(0
x.(0=*>
(4-102)
d 2 c(
**<*)
Then
t)
~ ~^m
dt*
the state equations are represented by the vector-matrix equation of Eq. (4-95)
"010
with
-2
(4-103)
-5
-1
and
B=
The output equation
(4-104)
is
c(t)
4.7
= Xl (t)
(4-105)
when
canonical form. It
tem with
is
and
are given
is
by Eqs.
(4-96)
single input
and
(see
Theorem
4-1.
given by
4*j&
= Ax(0 + Brit)
(4-106)
110
where x(t)
is
an n
1 state vector,
r(t)
S
is
Chap. 4
Dynamic Systems
State-Variable Characterization of
a scalar
an n
AB A 2 B
[B
n coefficient matrix,
matrix
an n X 1
input. If the
Anl B]
...
(4-107)
= Px(r)
(4-108)
or
x = P-'yCO
(4-109)
y(0
=A
y(0
B,r(0
(4-110)
where
...
...
1
1
...
(4-111)
A,
-a
and
(4-112)
is
given by
Pt
PA
t
(4-113)
_P A"1
where
P,
1][B
[0
AB A B
2
A"- B]]
(4-114)
Proof: Let
Xi(f)
x(0
(4-115)
_*.(0-
Sec. 4.7
>i(0'
y(0
(4-116)
U.(0.
and
Pll
Pl2
Pin
Pi
PlZ
Pin
P.
P=
(4-117)
Pnl
Pnl
Pnn.
where
P =
<
Pa
[Pa
P in ]
1, 2,
(4-118)
tim e
um
Fn!f "rf
fcqs.
(4-110)
,
and
J ^"
1
pJ=
il
+ p 12 x
(t)
2 (t)
...+ Pu xn(t)
n b th sMeS f the
last e 4 uation
and
=y
TSre^
2 (t)
P,x(0
^^
"
+ p.BKO
P,Ax(0
fUnCti n
(4-120)
0nly
in
'
with
P AB =
t
in view of
(4-111),
M*)
Eq
=P
Mt)
=y
3 (t)
P,A 2x(?)
(4 _ 12 i)
(4-122)
0.
with
P,A-*B
0.
= >'.(0 = PiA"-'x(0
(4-123)
we have
Pt
P.A
y(f)
Px(0
x(0
PjA"-
or
(4-124)
P,
P.A
P=
(4-125)
.Pi A"-'.
112
and P, should
Chap. 4
Dynamic Systems
condition
satisfy the
PjB
= P,AB =
P,A"- 2 B
(4-126)
Comparing Eq.
Px(0
PAx(f)
we
A,
we
get
PBr(0
(4-127)
obtain
= PAP
(4-128)
'
and
B = PB
(4-129)
P,B
'0'
'
P AB
t
PB =
__
_P,A"- B.
1
Since Pj
is
an
is
_1
P,[B
Thus P!
(4-130)
AB A 2 B
A"-B]
...
[0
...
(4-131)
1]
obtained as
P,=[0
...
AB A B
2
1][B
...
A-^B]-
(4-132)
= [0
...
lJS"
S = [B AB A B ...
1
2
A""'B] is nonsingular. This is the
the matrix
condition of complete state controllability. Once P! is determined from Eq.
(4-132), the transformation matrix P is given by Eq. (4-125).
if
Example
4-8
"1
-1
(4-133)
-1
It is
desired to transform the state equation into the phase-variable canonical form.
[B
AB]
ri
o~
(4-134)
'
-1
is
Therefore, Pi
of
S -1
that
is
may
obtained as a
is,
P,=[l
Using Eq.
-1]
(4-135)
(4-125),
(4-136)
_PjA_
Sec. 4.7
Thus
"0
=PAP'
A,
(4-137)
1
= PB = ro
B!
(4-138)
Given the
4-9
differential
^^ +
equation
~dk + ~dT +
L
2c
^-^dr +
2r(t)
(4-139)
it is desired to represent
the equation by three state equations. Since
the right side of
the state equations cannot include any
derivatives of the input r(t), it is necessary
to
include /(?) when defining the state variables.
Let us rewrite Eq. (4-139) as
d3c(0
dr(t)
,d*c(t)
,,
dc(t)
,
(4-140)
The
now
defined as
*i(0
= c(t)
(4-141)
(0=^
x2
(4-142)
(4-143)
Using these
last three
dt
=**(
^
dx 3 (t)
~3J~
In general,
dt-
+
'
it
= *.(0+r(0
.
= ~ 2x iW ~
,
(4-144)
x 2(0
5x 3 (t)
dr->
+a"-
~dt
*.^ +
dt"
+
...
a c
3r(f)
differential
equation
+ *._
^ +M0
dt
(4-145)
114
State-Variable Characterization of
Chap. 4
Dynamic Systems
- V(0
x,(t)
x 2 (t)
= ^f>-h
xAt)
= **M -
c(t)
r{t)
h 2 r{t)
(4-146)
*(0
^^-A
.-.'(0
where
>h
= a,6
= {b a b
= (* a b - * A -
A,
h2
bj,
(b
ab
flj/71
a_ x h x
aJh
(4-147)
a_ 2 h 2
-a
2 hn
_2
aj\ n _ x
differential
equa-
=x
-^f>
x 2 (r)
3 (t)
MO
h 2 r(t)
(4-148)
^M =
^1 =
xm (t)
+ h^r(0
-ax,(t)
is
a.,x 2 {t)
...
fl 2
x._,(0
x,(0
first
a iX {t)
+ hr(0
r{t)
4-9,
a, = 5
b =
b = 2
a =
6, =
b =
a = 2
= 6; tfj^o =
h = (b a b Vi =
a h ajx =
hi = (b a b
we have
/j,
When we
and
(4-147),
we have
Sec. 4.8
115
the same results for the state variables and the state equations as obtained in
Example 4-9.
The disadvantage with the method of Eqs. (4-146), (4-147), and (4-148)
is that these equations are difficult and impractical to memorize. It is not ex-
pected that one will always have these equations available for reference.
ever,
we
How-
function.
4.8
We
is
Eq. (3-16) gives the matrix transfer function relation for a multivariate
^
c(/)
= Ax(0 +
=
Dx(t)
is
Br(?)
(4-150)
+ Er(/)
(4-151)
where
=n
r (0 = P
=q
c
x(/)
X
X
input vector
(t)
output vector
state vector
we have
X(s)
(si
- A)" >x(0+) +
(4-151)
(si
~ A)-'BR(s)
is
Q = DX(j) + ER(s)
Substituting Eq. (4-152) into Eq. (4-153),
C(s)
D(sl
A)-x(0+)
(4-1 52)
(4-153)
we have
D(jI
- A)-'BR(5) + ER(s)
(4-154)
Since the definition of transfer function requires that the initial conditions be
set to zero, x(0+)
0; thus Eq. (4-154) becomes
C(s)
is
defined as
which
is
a q
matrix (si
x p
A)
is
matrix.
Of
(4-1 55)
(4-156)
nonsingular.
116
State-Variable Characterization of
Example 4-10
Chap. 4
Dynamic Systems
is
equations
+4
dt*
^+^ +
2
The
These
state variables
equations, as
no
+2c 2
c1
dt
(4-157)
iCl
dt
=/- 2
(4-158)
x2
= dci
-^
(4-160)
x3
(4-161)
(4-159)
Ci
c2
differential
particular reasons for the definitions are given other than that these
Now equating the first term of each of the equations of Eqs. (4-157)
the rest of the terms
(4-161),
we
dxi
and (4-158) to
in matrix
form:
xi
dt
dx 2
0-4
dt
ri
x2
(4-162)
1
Li- z J
dx 3
-1
ldt_
-2
-1
pi~|
ri
01
_C2.
_o
1_
x3
~Xi~
Dx
x2
(4-163)
_*3_
To
determine the transfer function matrix of the system using the state-variable
we substitute the A, B, D, and E matrices into Eq. (4-156). First, we form
formulation,
A),
-1
~s
(sI-A)=
The determinant of
(si
-3
+4
s3
+6s 2 +
(4-164)
2_
1]
A) is
|jI-A| =
(4-165)
Thus
~s z
(.51
The
A)
-| jI _ A
GCs)
is,
in
+6s +
-3
-(*
tllis
case,
= T>(sl -
+2
s(s + 2)
-(s + 1)
11
4)
f4-1
3i
fifi^
s(s+4)_
A)-'B
"
~ S* +
"
6s 2
+ Us +
s+2
_-(
(4-167)
3
!)
i(5
+ 4)J
conditions.
The
resulting transformed
and Eigenvectors
Sec. 4.9
117
+ 4)
+
we
sis
_ s
-3
nrc,(j)-i
+ 2j[_C
= VR^s)'
(4-168)
\_R 2 (s)_
2 is)j
obtain
C(.r)
G(j)R(i)
(4-169)
where
+ 4)
+1
Sis
Gis)
_ s
4.9
is
-3
(4-170)
+ 2_
obtained
carried out.
The
linear
trans-
is
equation
d"~ c u n d~ 2 c
"
ai
l
df
dF^
df-
an -\
dc
+ac
(4-171)
df
'df
By
p
P
Eq. (4-171)
is
as
Mi
A;
1,2, ... ,m
written
a 2 pn
=
Then the
is
setting the
operator p
The
ib P"
*,/>""'
+ a^^ + ajc
+ 1?-'P + b)r
..
which
is
+ a^"- +
1
a 2 s"- 2
G(s)
{S)
o-,J
an
(4-174)
is
- fro*" + M"" +
R(s) ~ s + a^ +
1
<&)
(4-173)
defined as
is
From
(4-172)
dt
is
+ b . iS + b
+ a _ lS + a
...
.
u
(4
'
175)
to zero.
G( , )
we can
= D adj(5l-A) B + E
si A
I
D[adj(sl
A)]B
UI-AI
jl
-A
(4-176)
|
118
/ State-Variable Characterization of
Dynamic Systems
Chap. 4
we
get
which
is
*I
A =
(4-177)
Eigenvalues
The
It is interesting
row of
the last
in
the
A matrix.
A = P"'AP
(4-178)
is
proved by writing
si- A = ^I-P'AP
(4-179)
or
si- A
The
characteristic equation of
\sl-
= ^""P-P-'AP
(4-180)
A is
AlHsP-'P-P-'API
(4-181)
=
Since the determinant of a product
|P-'(jI- A)P|
equal to the product of the determinants,
is
UI-
A|
= ]P-MI^I- A||P|
= |*I-A|
Eigenvectors
The n X
vector
p,
which
satisfies
a,I
where
X, is
the eigenvalue X t
Illustrative
- A)p, =
is
(4-183)
examples of
how
A associated with
4.10
Diagonalization of the
X2
located on the main diagonal; then the
x ,
A matrix is
X,
all
that if
assumed
to
A is
be
a diagonal
distinct, are
be
Sec. 4.10
Diagonalization of the
Matrix
119
its
x(0
= Ax(0 +
Bu(f)
(4-184)
y(0
with
Ay(r)
ru(f)
(4-186)
A.
...
23
...
(n
is
n)
(4-187)
...
This transformation
also
is
state
it is
P 'AP
(4-188)
and
P'B
(n
r)
(4-189)
P=
where
p, (/
eigenvalue k
is
P 3 ...pj
P2
= Ap,-
i=
1,2,.
is,
(4-190)
is
Now
[Pi
This
We show
,n
written
(4-191)
UiPi
A 2p 2
Ap]
= [Ap,
= A[p,
p2
p2
Ap 2
Ap]
(4-192)
p]
or
[Pi
Therefore,
if
we
P2
P]A
A[p,
P.J
(4-193)
let
P=
[Pi
P2
P.]
(4-194)
120
Chap. 4
Dynamic Systems
State-Variable Characterization of
PA = AP
(4-195)
or
A = P-AP
which
is
If the matrix
the
(4-196)
X\
X\
...
matrix,
K
XI
(4-197)
X~\
where X u X 2
Since
of A,
we
it
shall
eigenvector of
show
that
is
Xi=
associated with
1,2,...,
n.
Let
Pn
(4-198)
P,
Pin.
be the
rth eigenvector
of A Then
.
U,i
A)p, ==
(4-199)
or
~x
-1
x
Pn
-1
x<
Pn
-1
(4-200)
-1
_0
a n -i
a-!
a- 3
Cl\_
-Pin
XiPn
XiPn
Pn =
Pn =
(4-201)
Pin=
+ a )p =
XtPi,-l
fl//i
a n-\Pn
+ (X
tm
Sec. 4.10
Now
we
Diagonalization of the
arbitrarily let
pn
Then Eq.
1.
=
=
Pn
Pn
Matrix
121
(4-201) gives
X,
tf
(4-202)
Pt,n-1
xr
Pin
Ar
Example
4-11
p,
rth
in Eq. (4-197).
equation
is satisfied.
A=
on
(4-203)
-11
-6_
"1
P =
The canonical-form
state
1"
X\
"I
A3
-2
-3
_Aj
Ai
A3
equation
is
(4-204)
"-1
A =piAP
"a,
-2
A2
(4-205)
A3.
Example 4-12
-r
A=
11
11
5j
is
(4-206)
A is not
pn
Pi
Then
pi
must
= Pn
(4-207)
satisfy
a,i
- 1^1
==
(4-208)
122
State-Variable Characterization of
Dynamic Systems
Chap. 4
or
-1
~X 1
6
A,
_6
The
last
"
Pu
-6
A -5_
11
11
(4-209)
Pi\
J>31.
Pu Pzi + P3l =0
+ 10p 21 6p 31 =
6p n + 11^21 6/31 =0
= and/?n = p 3i Therefore, we
(4-210)
6pu
can letpn
= /> 3 =
i
1,
and
get
(4-211)
Pi
-1
"A 2
A2
"
-6
ll
(4-212)
P11
A 2 -5.
11
_/>32.
or
(4-213)
22
32
4.
Thus
P2
(4-214)
4
Finally, for the eigenvector p 3 ,
we have
-1
~X 3
A3
_6
P2 3
A 3 -5.
-P33.
(4-215)
3/> 13 Pu + P33 =0
6p 13 + Sp 23 - 6p 33 =
6pi + Up 2 8/? 33 =0
if
we
arbitrarily let
i3
1,
(4-216)
Now
Pl3
-6
11
11
or
"
6 and
p 33 =9.
Therefore,
p3
The matrix P
is
now
(4-217)
given by
"1
P =
[Pi
P2
P 3]
_1
9_
(4-218)
Sec. 4.11
It is
easy to
show
123
that
"A,
A=P
AP
on
-2
A2
(4-219)
-3_
A3
_0
4.11.
matrix
is
A = P'AP
(4-220)
ri)
such that the matrix A is almost a diagonal matrix. The matrix A is called the
Jordan canonical form. Typical Jordan canonical forms are shown in the following examples
[A,
A,
A=
A,
(4-221)
A2
_0
"A:
A3
1
A,
A=
A2
(4-222)
A3
_0
The Jordan canonical form
1.
A4
matrix.
2.
3.
Some of
4.
5.
The Is, together with the eigenvalues, form typical blocks which
are called Jordan blocks. In Eqs. (4-221) and (4-222) the Jordan
blocks are enclosed by dotted lines.
When
its
is
6.
main diagonal of
are zero.
the elements immediately above the multiple-ordered
eigenvalues on the main diagonal are Is, such as the cases illustrated
by Eqs. (4-221) and (4-222).
the nonsymmetrical
matrix has multiple-order eigenvalues,
eigenvectors are not linearly independent. For an n X n A, there
only r
(r
<
7.
the
r.
124
State-Variable Characterization of
Dynamic Systems
The matrix P
sume that A has q
is
Chap. 4
distinct eigenvalues
among n
eigenvalues. In the
first
as-
place,
manner from
(A I
f
where
A)p,
A,
The
mined by
(4-223)
2,
q.
A,
...
(m X m)
...
(4-224)
A,-
Xj_
.0
Then
Vz
0"
...
Ay
[Pi
must hold:
A[p,
PJ
o
kj
Lo
AyPi
Pm-i
The
vectors Pi, p 2 ,
also
be written
+ AyP
+ AyP
(4-225)
Ay.
P2
PJ
or
p2
A,p m
= Ap,
= Ap
= Ap
(4-226)
= Ap m
- A)Pl =
(Ayl - A)P = -P,
(Ayl - A)p = -P
(Ayl
(Ayl
Example
4-13
A)p m
(4-227)
= P m _l
A=
-5"
4_
(4-228)
Sec. 4.11
the determinant of Al
IAI
125
A is
-Al
-6
-1
-3
-2
- 4A 2 +
A3
5A
2
(4-229)
-2 A -4
(A
2)(A
l) 2
Therefore,
To
such
find the
mined from
(A,I
A)p,
(4-230)
Thus
-6
-1
-3
Setting
pn
5"
-2
/>n
-2
^21
-2_,
(4-231)
_/>31_
equation gives
p zl
and pi
There-
-2.
fore,
2"
(4-232)
Pi
to Eq.
(4-227).
(A 2 I
- A)p 2 =
(A 2 I
(4-233)
and
= -p
A)p 3
(4-234)
-1
-3
Setting
p 12
arbitrarily,
-6
5"
P\i
-2
-2
Pn
-3_
(4-235)
_/>3 2_
= ^ andp 32
wehave/? 22
1"
(4-236)
Pz
-6
-1
-3
- 5" Pl3
2
P23
3_ _P33_
-r
3
7
L
(4-237)
7J
Pl3
P3
Pl3
1"
-n
(4-238)
_46
_P33_
Thus
-7
-t
-n
-li
(4-239)
126
State-Variable Characterization of
Dynamic Systems
is
Chap. 4
now
obtained as
~2
A = P'AP
0"
1
Note
(4-240)
_0
1_
is
4.12
State Diagram
The signal flow graph discussed in Section 3.5 applies only to algebraic equations.
In this section we introduce the methods of the state diagram, which represents
an extension of the signal flow graph to portray
The important
equations.
close relationship
among
all
and
diagram
that
is
differential
it
forms a
lowing
state equations
state
diagram
is
constructed fol-
the rules of the signal flow graph. Therefore, the state diagram
may
it is
The fundamental
constant
is
tion
x 2 (t)
where a
is
a constant. If a
lies
a Xl (t)
(4-241)
late
Eq. (4-241)
of a considered
is
x 2 (t)
*i(0
xAt)
x 2 )
<z>
x 2 (t) = ax
Fig. 4-5.
(t)
0<a<\
Analog-computer block-diagram
symbol of a potentiometer.
x 2 {t)=ax
Fig. 4-6.
{t)
of an operational amplifier.
State Diagram
Sec. 4.12
127
Algebraic sum of two or more variables. The algebraic sum of two or more
machine variables may be obtained by means of the operational amplifier.
Amplification may be accompanied by algebraic sum. For example, Fig. 4-7
*i(0
x 2 (t)
*-
x,(t)
x 3 (t)
x 4 (0 = a, x l
(t)
+ a 2 x 2 (t) + a } x 3 (t)
summer.
analog computer block diagram of a summing operational amwhich portrays the following equation
illustrates the
plifier
x 4 (t)
Integration.
The
a^At)
a 2 x 2 {t)
+a
x 3 (t)
(4-242)
on an analog computer
is
initial
condition Xi0
given at
is
the
and x 2 (t)
is
x,0)
f ax 2 {x)
>
dz
JCi(f )
<
(4-243)
to
The block diagram symbol of the integrator is shown in Fig. 4-8. The integrator
can also serve simultaneously as a summing and amplification device.
*i('o)
x 2 (t)
-- x,(t)
*i( f )=
Fig. 4-8.
ax 2 (r)dT +
*,(?<))
integrator.
We shall now show that these analog computer operations can be portrayed
by
state
X (s) = aX^s)
2
The
is
shown
in Fig. 4-9.
we
take the
(4-244)
128
State-Variable Characterization of
Chap. 4
Dynamic Systems
*i(0
x,(0
x 2 (t)
^i(s)
X2 (s)
Fig.
4-9.
x 4 (t)
X2 (s)
XAs)
Signal-flow-graph
= aix\{t)
+ a 2 xz(t) + ajXi(i) orX4 (s) =
aiX^s) + a 2 X 2 (s) + ai X3 (s).
x 2 (t)
of
representation
axi(t) or
*2(0
X (s) =
2
resentation of Xi(i)
aX\(s).
is
X (s) =
4
The
,*,(*)
a 2 X2 (s)
a s X3 (s)
is
(4-245)
shown
in Fig. 4-10.
important to note that the variables in the signal flow graphs of Figs. 4-9
and 4-10 may be in the time domain or the Laplace transform domain. Since
the branch gains are constants in these cases, the equations are algebraic in
It is
both domains.
sides
of Eq. (4-243). In this case the transform operation is necessary, since the signalflow-graph algebra does not handle integration in the time domain. We have
X 2 (T)dT\
x 2 {x) dx
+ *i(*o)
x,(t )
x 2 (t) dx
(4-246)
Jo
x 2 (r) dx
X^s)
It
^^ + ^5)
is
(4-247)
is
is
now
shown in Fig.
shown in Fig. 4-12.
flow graph as
algebraic
4-11.
An
>
(4-247)
to
X^s)
>
signal
Sec. 4.12
State Diagram /
129
*i('o)
*i('o)
<
as- 1
>
X2 is)
=
[aX 2 (s)ls]
[xiOo)/*]
<>
Xi
(s)
X2 (s)
Xds)
Xi(s)
resentation of Xi(s)
[aX 2 (s)/s]
[xi(.to)ls].
may
let
state
directly
2.
state
function.
4.
The
a digital computer.
5.
The
may be
6.
The
from the
state
diagram.
7.
The
state equations
from the
The
From
details
state
diagram.
When
Diagram
130
State-Variable Characterization of
Chap. 4
Dynamic Systems
equation
d"c
d"
dc
._
(4-248)
we rearrange
the equa-
tion to read
l
d"- c
d"c _
dt"
o
R
O
s"C
n~l
o
C
s"- 2
dc
-a. iw -ac-rr
<
(4-249)
o
c
(a)
(b)
("-!)/"
(n-2>,
(1) (f
c ( ? o)
(c)
differential
equation of Eq.
Sec. 4.12
State Diagram /
131
Now
as
the variables
shown
r, c,
W
is
'=1.2,...,b
(4-250)
c (1> , c (2) ,
(n>
('
(,'>
When
the input
is
on the
not so straightforward.
We
is
state
diagram
directly
show later that, in general, it is more confunction from the differential equation first and
shall
Example 4-14
differential equation:
g + 3| +
2c
=r
(4-251)
2c
rest
of the terms,
we have
dT^- 2c - 3 dl
_
dc
(4-252)
Following the procedure outlined above, the state diagram of the system
is
shown
in
Fig. 4-14.
c (1) a +)
9*o+)
"
to
(4-251).
was mentioned earlier that the state diagram is essentially a block diaprogramming of an analog computer, except for the phase reversal
through amplification and integration.
It
gram
for the
132
Example 4-15
Chap. 4
Dynamic Systems
State-Variable Characterization of
An
(4-251)
may be somewhat
may be necessary.
from shown,
and time
since amplitude
scaling
different
Fig. 4-15.
Eq. (4-251).
to Digital
Computer Simulation
FORTRAN
same purpose
is
as an analog
practically eliminated.
typical
CSMP
The
CSMP
Mathematical Equations
c
a x
1
a 2x 2
x.
Xl
*,(t) dx
f'
Example 4-16
From
x,(0)
Statements
C=
A\
Y=
JT1/2.
XI
XI
A2 * X2
INTGRL(Z2,
JT10)
written
dt
(4-253)
c=\cdt
c = r - 3c - 1c
(4-254)
f t
(4-255)
Sec. 4.12
State Diagram /
133
c
c
r
=C
= CI
= C2
=R
= CO
= CIO
c(0)
<?(0)
CSMP representation
of the system
is
given as follows:
(4-256)
CI
We
(4-257)
(4-258)
have shown
With the state diagram, the equivalence of the matrix inverse operation is
out by use of the signal-fiow-graph formula.
The state transition equation in the Laplace transform domain is
X(s)
(jI
A)- !x(r J)
(jI
A)" 'BR(i)
>
A).
carried
(4-259)
Therefore, the last equation can be written directly from the state diagram by
Example 4-17
and Rj(s),j
1, 2,
1, 2,
n, as
and
Consider the state diagram of Fig. 4-14. The outputs of the intediagram is
redrawn as shown in Fig. 4-16.
Xl (s)
=J
X (s) =
2
"'
(1
-2s~
2 (*o)>
3;rl)
state
anc*
>-
diagram
X (s)
t
and
+ _L
v.(t*\
*l('o )
XlOZ)
r -2
*fr)
(4-260)
(4-261)
34
State-Variable Characterization of
Chap. 4
Dynamic Systems
where
A=
+3s~ +2s~ 2
(4-262)
After simplification, Eqs. (4-260) and (4-261) are presented in matrix form:
xiin)
(j
Xz(s)
IX*
2)
-2
(s
L(j
The
>
R(s)
(4-263)
iXj
2)J
is
2)
xAn)
l)(s
r{t) is
at
Then
the
- (-1)
"
The
u,{t
2e~
Mt).
2e
Example
g-c-o)
e -2d-r
' o)
g-c-'o)
-|_
+ e-
<'-'>
this result
x 2 (t$)_
(4-266)
2 <'-'>"
?>
and
> 0.
is
4-18
setting x^{t J)
C(s)
t_
R(s)
s2
The
xiCt)
2e~ 2( '~' )
to Transfer Function
transfer function
diagram by
- c"
t )
(4-265)
is
'
p-2(t-lo)
(l-lo)
'
*'
-( '~''
t u s {t
e -2 + 2e~ 2(t_
(,- ' )
(4-264)
t>t
<
(jTTi)
'
t>t
to)
3s
and
Xi(t J) == 0. Therefore,
+ 3s +
(4-267)
is
(4-268)
When the state diagram of a system is already given, the state equations
and the output equations can be obtained directly from the state diagram by
use of the gain formula.
Some
xi(t
2,
n,
and the
initial
states,
'
1 2,
n, are
domain
The
Sec. 4.12
State Diagram /
135
The left side of the state equation contains the first-order time derivative
of the state variable x (t). The right side of the equation contains the state
t
variables
state variables in
Fig. 4-17.
initial states
and
r as
input nodes,
the integra-
written directly:
^ = -2x
(4-269)
dx 2
dt
Example 4-19
3x 2
As another example of
or
(a)
136
State-Variable Characterization of
Dynamic Systems
Chap. 4
O
r
*3
(b)
diagram
and
inte-
grators eliminated.
and the integrators being eliminated. Notice that in this case the state diagram in Fig.
4-18(b) still contains a loop. Applying the gain formula to the diagram of Fig. 4-18(b)
withii,;c 2 and x 3 as the output node variables and r,x\,x 2 and x 3 as the input
,
dxi
Xi
dt
dx 2
dt
(a 2 + a
+aa
1
O 2
3)
aa a 3
dx 3
(4-270)
*3
dt
4.13
x2
Up
this
methods of characterizing a
linear system
have
been presented.
It will
may
system
equations. It
from
shown in
block diagram
is
drawn
as
with the differential equation of a system, one can get to the solution by use of
Sec. 4.13
Differential
Dynamic
equations
equations
137
'
State
transition
equation
^
Transfer
function
State
diagram
Fig.
One
ways of decomposing a transfer function direct decomposicascade decomposition, and parallel decomposition. Each of these three
schemes of decomposition has its own advantage and is best suited for a par-
tion,
ticular situation.
Direct Decomposition
C(s)
R(s)
The
objective
is
s2
b s*
+as+a
+bs+b
(4-271)
state equations.
The
fol-
of
s.
This
power
is
j.
the numerator
2.
C(s)
R(s)
3.
a
b
+ a s + a s~
s~
+ b,s' + bh-*~
'
X{s)
X{s)
(4-272)
138
State-Variable Characterization of
Chap. 4
Dynamic Systems
From
Eq.
to each other,
=
R(s) =
C(s)
4.
a,s~
(fl
(b
+ *,*-
a 2 s- 2 )X(s)
(4-273)
b t s-*)X(s)
(4-274)
they must
first
X{s)
However,
and must be rearranged.
by b and writing X(s) in terms of
left side
relation. It is ap-
we have
b2
ba
(4-275)
'X(s)
Eqs. (4-273)
As
usual, the state variables are defined as the outputs of the integra-
The
diagram.
state
tors.
C(s)
R(s)
direct
decomposition.
Xl
dt
dx 2
.dt
-b 2
(4-276)
-6i
b
equations are
x2
J
The output equation is obtained from Fig. 4-20 by applying the gain formula with c(t) as the output node and x^t), x 2 {t), and r(t) as the input nodes.
(4 - 277)
Sec. 4.13
Decomposition
of Transfer Functions /
139
Cascade Decomposition
Cascade decomposition
may
the factored form. Consider that the transfer function of Eq. (4-271)
is in
may be
factored in the following form (of course, there are other possible combinations
of factoring)
gi
R(s)
where z u
z 2 ,p u
and p 2
= ^lIilJi
b s + p s + p
Then
(4-278)
it is
two
first-order
two
R(.s)
~Pi
Fig. 4-21. State
decomposition.
integrators
on the
state
state variables.
The
state
dx-i
-Pi
dt
Pi
ctx^
_*oJ
is
(z 2
p 2 )x +(z,
l
- Pi)x +
2
facilitates the
(4-280)
This
(4-279)
*2
-Pi
Ldt \
The output equation
a
b
on the
when
and zeros of
state diagram.
the poles
and zeros
are varied.
Parallel
Decomposition
When
is
in factored form,
it
is
Cjs)_
R(s)
where P(s)
is
Pis)
+ Pl )(s + p
(s
a polynomial of order
less
than
2.
(4-281)
2)
We
140
Chap. 4
Dynamic Systems
State-Variable Characterization of
p2
is
written
where
The
the state
and
K,
Kr
C(s)
R(s)
+ Pi
+p
(4-282)
2
are constants.
state
shown
in Fig. 4-22.
The
state equations
written
dxC
dt
P\
(4-283)
dx 2
X%
-Pi
\_dt _
X!(f +)
R(s)
parallel
decomposition.
is
One of
[K,
Therefore,
we can
A matrix is always
diagonalization of the
When
(4-284)
2]
is
a diagonal matrix.
may be
A matrix.
that the state diagram, as obtained through the parallel decomposition, contain the minimum number of integrators. To further clarify the point just made,
its
partial-fraction expansion:
Sec. 4.14
an
R(s)
2s 2
(s
6s
iy(s
+5
+ 2)
(s
141
+ s+l
iy
(4-285)
Note that the transfer function is of the third order, and although the total order
of the terms on the right side of Eq. (4-285) is four, only three integrators should
be used in the state diagram. The state diagram for the system is drawn as shown
in Fig. 4-23. The minimum number of three integrators are used, with one in-
paralled decomposition.
state equations
written
-1
*1
0-1
x2
dt
dx 2
dt
dx 3
0-2
dt
Therefore, the
4.14
A matrix is
x3
(4-286)
When
form
the
it
it
putation,
A = P"'AP
Let us assume that
jco.
is
x 2 and has
is
eigenvalues A,
given by
(4-287)
+ jco
and X 2
142
State-Variable Characterization of
Dynamic Systems
Chap. 4
CO
A=
(4-288)
CO
The elements of the P matrix may be determined by brute force using Eqs.
(4-287) and (4-288). If A has m real distinct eigenvalues in X u X 2
X m and
n sets of complex-conjugate eigenvalues in X, = a jcoi, i = 1 2,
,
,
is
given by
A,
..
x2
A=
Am
..
..
o
(4-289)
..
..
..
where
(4-290)
On
\CO,
If they'th
Tj
r,
r,
of multiplicity m, then
is
is
written
...
(m X
blocks)
(4-291)
ooo
where
"
Oj
COj~
i-COj
CTj]
(4-292)
ri
oi
Lo
lj
(4-293)
in Eq. (4-289)
is
easily
modified for
real
and multiple-
decoupling of the states from the standpoint of the state diagram, it still has
the components of the eigenvalues as its matrix elements.
To determine the transformation matrix P for the matrix A of Eq. (4-288),
we
let
P=
where p and p 2 are 2 x
(
[Pi
vectors.
P2]
(4-294)
pj
[Pl
Equation (4-287)
co
-co
A[p,
_
is
p2 ]
written
(4-295)
Sec. 4.14
or
cop = Ap!
toPi + cq>2 = Ap
ffPi
143
(4-296)
(4-297)
-coll ~pr
col
rf_ J>2_
"A
0~
_0
A_
Pi
(4-298)
where
denotes a 2
_P 2
2 identity matrix.
Let qi and q 2 denote the eigenvectors that are associated with the two
complex-conjugate eigenvalues, X 1
a
jco and X 2
a ja>, respectively.
(^+yco)q
- jco)q
(<r
=Aq
= Aq
satisfy
(4-299)
(4-300)
Let
=
=
q,
q2
Then
<X,
(4-301)
(4-302)
become
(4-303)
(<7
2)
Equating the
real
_coI
last
and
a\
_ col
Comparing Eq.
coT "2~
0" "2~
LP 2 J
_0
A_ LpzJ
<rl_
The
we have
P=[P.
(4-305)
A_ LpJ
LpJ
ctI
01 rii
"A
col "i
ctI
(4-304)
P2]
[i
(4-306)
the identity
(4-307)
Pi]
P is formed
and imaginary components of the eigenvector of A associated
by taking the
with Xi
Example 4-20
real
+ jco.
Consider the state equation
i = Ax
+ Br
(4-308)
where
1"
-2
The eigenvalues ofA are A
B=
-2
qi
q2
_-i -J.
or
qi
a! +/Pi
q2
= a 2 +/P2
+ J -1
144
State-Variable Characterization of
Dynamic Systems
Chap. 4
Therefore,
P =
A=
[*i
Pt]
P-!AP
_-i
(4-309)
i-
= r-i
i~|
-i
~i_
(4-310)
T = piB =
The
Then
is
(4-311)
is
transformed to
= Ay + Tr
given by
= eM =
~ cos
Controllability of Linear
sin i
(4-313)
e~<
sin
4.15
(4-312)
cos
Systems
observability introduced
first
by Kalman 24
The conditions on
controllability
tivariable systems.
that the system be controllable and/or observable in order to achieve the control objectives.
this section
we
These points
will
be
of controllability.
The concept of
diagram of
Fig. 4-24.
The process
is
State x(t)
Control u(0
G can
u(t). Intuitively, it is
simple
state, the
trollable.
system
if
is
said to
Sec. 4.15
Controllability of Linear
Systems
145
?*i (>o+)
?*i('o+)
u(t)
is
As a
state
i/(0 affects
is
tf
ta
initial state
by any control
x 2 (t
u(t).
it
to a desired state
would
x 2 (tf)
said to be uncontrollable.
The concept of controllability given above refers to the states and is sometimes referred to as the state controllability. Controllability can also be defined
for the outputs of a system, so there
and output
is
controllability.
dynamic equations
is
x(0
c(0
=
=
+ Bu(0
Dx(0 + Eu(0
Ax(0
(4-314)
(4-315)
where
=n x
u(0 = r X
c(0 = p x
A=n X
B=n X
D =p x
E =p x
x(0
state vector
input vector
output vector
n coefficient matrix
r coefficient
matrix
n coefficient matrix
r coefficient
matrix
controllable.
>
146
Stale-Variable Characterization of
Dynamic Systems
Chap. 4
way of testing
state controllability.
Theorem 4-2. For the system described by the state equation of Eq. (4-314)
be completely state controllable, it is necessary and sufficient that the following
n X nr matrix has a rank of n:
to
AB A
[B
B.
A"-'B]
(4-316)
Since the matrices A and B are involved, sometimes we say that the pair
is controllable, which implies that S is of rank n.
[A, B]
Proof:
The
<j>(?
)x(t )
is
>
for
>
tQ
finite
f >
is
x('o)
t)Bu(t) dx
(4-317)
x(t f )
=~
f"
Then Eq.
0.
Mo -
final
(4-317) gives
(4-318)
t)Bu(t) dx
From
A*
Then
some
state for
ftt
to
= 2
a* m A
for any
(4-319)
<K0
S A"t
*=o k}
'=
icl
(4-320)
Zj TT
Kk=
2-1
m=0
&km-
or
ftf)
Thus
<(?)
n-l
2 a km U
k=0
(4-321)
A"
(4-322)
oo
"t A"
m-0
,k
ft!
<K0
= - 2 Am B
f"
a m (f
t)u(t)
</t
we have
(4-323)
Let
U m = r m 0o-^)u(T)^T
J
Then Eq.
(4-323)
becomes
x('o)
= - 2 A m BU m
771
which
is
(4-324)
to
(4-325)
Sec. 4.1 5
Controllability of Linear
x('o)
= -[ AB A B
= -SU
2
Systems
A""'B]U
147
(4-326)
where
U=
[U
U....U,.,]'
(4-327)
Equation (4-325) represents n equations with nr unknowns, and the conproblem may be interpreted as: Given any initial state x(t ), find
trollability
is
x(^)
S, solve
the system
linearly
Although the
by Theorem 4-2 is
not very easy to implement for multiple-input systhere are In columns in S, and there would be a large
quite straightforward,
it is
Even with r = 2,
number of possible combinations of n x n matrices. A practical way may be
to use one column of B at a time, each time giving an n X n matrix for S. However, failure to find an S with a rank for n this way does not mean that the system is uncontrollable, until all the columns of B are used. An easier way would
be to form the matrix SS', which is n X n then if SS' is nonsingular, S has
tems.
rank
n.
Example
4-21
Consider the system shown in Fig. 4-25, which was reasoned earlier to
be uncontrollable. Let us investigate the same problem using the
condition of Eq. (4-316).
The
rki(Oi
-2
dt
dt
iW
dx 2 (t)
I
-1
u(t)
(4-328)
x%{t)
S
which
is
Example 4-22
Determine the
[B
is
AB]
"1
-2"
_o
o.
(4-329)
e.
state equation
dxM
+
dx 2 Q)
.
From
dt
*i(0
dt
-1
"(0
Eq. (4-316),
[B
AB]
(4-331)
1
which
(4-330)
xi(0
is
is
148
State-Variable Characterization of
Dynamic Systems
Chap. 4
i(0
If the eigenvalues
A are
of
Ax(0
distinct
is
Bu(?)
(4-332)
n,
into a
such that
...
"A,
A,
P'AP =
(4-333)
0.
Let the
new
state variable
be
y
Then
P-'x
(4-334)
Ay
is
(4-335)
Tii
where
r = p-'B
(4-336)
The motivation
eigenvalues
eigenvalues can
be diagonalized.
Does
but whose
The
natural question
is:
is
no.
We
must not
lose
on
^
^
state equations:
a Xi (t)
+ bMO
(4-337)
= ax&)
b 2 u{t)
(4-338)
Controllability of Linear
Sec. 4.15
This system
is
ab
~b,
AB]
[B
mean
(4-339)
149
ab 2
bz
is
S
is
Systems
diagonal, and
is
controllable.
is
When A
is
a nonsingular matrix
A = P'AP.
The elements
X it X u X u
which transforms
is
in the other
all
Xi, three
A into
if
all
the matrix
P AP
four eigen-
a nonsingular
form
0~
A,
A has
is
'
(4-340)
A,
A2 J
self-explanatory.
Example 4-23
respectively,
"-2
-i_
B=
and
matrices are,
T
l_oJ
Let us check the controllability of the system by checking the rows of the matrix T.
It
A is
p=
_o
Therefore,
'B
-r r =
"1"
i_ _o_
_0_
_o
The transformed
state equation
(4-341)
is
~-2
0~
no
y(0
(4-342)
u(t)
_0_
row of
uncontrollable.
Example 4-24
is
is
x 2 (0
*s
uncontrollable,
-4
A=
-r
"0
B=
150
State-Variable Characterization of
Dynamic Systems
Chap. 4
Then
AB A
[B
-1
-41
8_
B]
(4-343)
1
Since
A2
2,
0"
(4-344)
1
1
2_
Then
01
[2
A = P'AP
(4-345)
_0
1_
T = P'B
Since the last
this
row of T
is
(4-346)
y3
is
uncontrollable. Since
x2
=y
3,
Example 4-25
state
equation
i(0
-l
x(0
u(t)
(4-347)
= [B
1"
AB]
(4-348)
1
which
is
nonsingular.
Let us
now
eigenvalues of
formation,
~
"1
J
=
-i]
AP
7
_o
"
/_
and
r=
p-'b
=
L2/J
Since
all
the rows of
is
controllable.
In general, when the eigenvalues are complex, which occurs quite frequently in
control systems,
it is
more
difficult to
Sec. 4.15
Controllability of Linear
Systems
151
use the modal form so that only real matrices are dealt with. In the present problem
A may
CO"
_a>
a_
r
(4-349)
.-1
o_
-r
_1
i_
P =
Then
r=p
'B
2"
1
-1Since the
modal form
lability is that
not
rows of
all the
are zeros.
Output Controllability 1 *
is
is
referred
controllable if every
by an unconstrained
sary nor sufficient for the existence of a solution of the problem of controlling
of output
output controllable
if
controllability.
system
is
said to be completely
(t f
t )
>0.
Theorem 4-3. Consider that an nth-order linear time-invariant system
described by the dynamic equations of Eqs. (4-314) and (4-315). The system
completely output controllable if and only if the p X (n
T)r matrix
is
is
T = [DB DAB DA 2 B.
is
of rank p. Or,
E]
(4-350)
is
similar to that of
Theorem
4-2.
dt 2
state controllability
gated.
DA nl B
'
d 2 c(t)
The
dc(t)
~dT
c{t)
_ du{i)
~ -dT
is
described
(4-351)
We shall show that the state controllability of the system depends upon how
x2
=
=
c
c
the
152
Dynamic Systems
State-Variable Characterization of
The
_-i
-2_
"
=
_*2_
Chap. 4
= X\
is
matrix
The system
singular.
From
_-l
i.
(4-354)
is
D=
[1
T = [DB DAB
is
AB]
0]
and
E]
E=
The output
0.
controllability
written
is
which
(4-353)
is
= B
S
which
(4-352)
-l
is
The
+
X%-
of rank
the
1,
-1
[1
(4-355)
0]
is
output con-
trollable.
Now
let
"
is
now
"0"
["1 +
-2. L*2_
Xi
+x
(4-357)
[B
AB]
-2,
ZJ
"0
is still
T = [DB DAB
is
(4-358)
nonsingular.
The system
which
(4-356)
_1_
_i
is
the
S
which
By
form
is
The system
.-1
_*2_
of rank
E]
[1
-1
0]
(4-359)
We
have demonstrated through this example that given a linear system, state
depends on how the state variables are defined. Of course, the output
controllability is directly dependent upon the assignment of the output variable. The
two types of controllability are not at all related to each other.
controllability
4.16
The concept of
tially,
a system
observability
is
is
completely observable
if
some of the outputs. In other words, it is often desirable to obtain information on the state variables from measurements of the outputs and the
inputs. If any one of the states cannot be observed from the measurements of
the outputs, the state is said to be unobservable, and the system is not completeaffects
ly observable,
or
is
the state
Sec. 4.16
?*2('0+)
153
?-M'o+)
u(t)
is
not observable.
is
described by the dynamic equations of Eqs. (4-3 14) and (4-315), the state \(t )
is said to be observable if given any input u(f), there exists a finite time t
tQ
f
>
<t <
<t<t
V=
[D'
A'D'
(A') 2
D\
(A')"- 1 D']
(4-360)
B(f>(t
)x(t
+D
<f>(?
we have
t)Bu(t) dx
Eu(?)
(4-361)
c(0
Making
D$(r
)x(t
(4-362)
c(0
m=
a m (0DA-x(r
(4-363)
154
Dynamic Systems
State-Variable Characterization of
Chap. 4
or
D
DA
DA
c(0
a I
(a I
(4-364)
xOo)
-,!)
DA"
Therefore, knowing the output
is
c(t)
and only
if
D
DA
DA
if
< <
t
f , x(t )
the matrix
(np
DA"
has rank
n.
Or
(A') 2
A'D'
[D'
(A')"-'D']
(4-365)
n.
Comparing Eq.
following observations
1.
the matrix
V=
has a rank of
w)
may
Controllability
made
be
[A', B'].
2.
Example 4-27
Consider the system shown in Fig. 4-26, which was earlier defined to
be unobservable. The dynamic equations of the system are written
directly
from the
state diagram.
'-2
~3"
0"
-1_ L*2_
(4-366)
_1_
~x{~
(4-367)
0]
[1
_*2_
L*2j
Therefore,
D=
[1
-2
A'D'
D'
0]
0"
-1
V=
Since
is
is
[D'
A'D]
unobservable.
"1
-2'
(4-368)
Sec. 4.16
Example 4-28
155
-1"
1.
Xl
(4-369)
L*2.
ci
(4-370)
-1
For the
test
of observability,
AD'
The
observability matrix
has a rank of
Lxil
we
evaluate
n ri
-ii
_-i
i- J)
i_
_-l
(4-371)
2_
becomes
smes
V
Since
Xl.
)'
2,
which
AD] =
the
is
"i
--l
0"
_o
-1
2_
number of
(4-372)
is
completely
observable.
Example 4-29
are defined.
a system depends on
how
on the
definition
of the state variables. Let the dynamic equations of the system be defined as in Eqs.
(4-352)
and
(4-353),
A=
D=[l
0]
-1
Then
V=
[D'
AD]
(4-373)
D=
-1
-2
[D'
AD] =
[l
(4-357).
1]
Then
V=
"1
']
which
is
singular.
is
-1
-1
unobservable, and
we have shown
one method of state variable assignment, Eqs. (4-352) and (4-353) yields a
system that is observable but not state controllable. On the other hand, if the dynamic
equations of Eqs. (4-356) and (4-357) are used, the system is completely state control-
(4-351),
investigate these
There are
phenomena
definite reasons
behind these
results,
and we
shall
of observability. If the matrix A has distinct eigencan be diagonalized as in Eq. (4-333). The new state variable is
Alternative definition
values,
it
P-'x
(4-374)
156
State-Variable Characterization of
Dynamic Systems
Chap. 4
are
=
+ Tu
Fy + Eu
F = DP
Ay
(4-375)
c ==
where
Then the system
is
(4-376)
(4-377)
The reason behind the above condition is that if the /th (j = 1,2, ... ,n)
column of F contains all zeros, the state variable ys will not appear in Eq.
(4-376) and is not related to the output z{i). Therefore, y, will be unobservable.
In general, the states that correspond to zero columns of
Example 4-30
above requires that the matrix D = [1 0] must not contain any zero columns. Since
the second column of D is indeed zero, the state x 2 is unobservable, and the system
is
4.17
Relationship
unobservable.
Among
Controllability, Observability,
Although controllability
and observability are concepts of modern control theory, they are closely
Let us focus our attention on the system considered in Examples 4-26 and
4-29. It
was demonstrated
state controllable or
which has an
_
s*
identical pole
is
+ 2s +
and zero
have
+ _
, 78
/oj
s+l
+ l)
= The following theorem gives
s
~(s
at s
We
(
<.<->
is
=
c(t) =
x(?)
Ax(t)
Dx(0
Bu(t)
(4-379)
(4-380)
Sec. 4.17
Let the
Relationship
A matrix
Among
state
A=
...
/]
ki
A3
...
X\
XI
X\
1n-l
A1n-l
2
DP. The
is
Tu(t)
(4-382)
(4-383)
by
Py(f)
(4-384)
is
a, is
*#&)
P,
transformed into
is
and
157
(4-381)
Fy(0
].-
MO =
where
ln-1
A3
Ay(?)
x(r)
Since
A?
and Functions
Vandermonde matrix
c (0
where F
y(0
where
be diagonalized by an n
.Ai
The new
Controllability, Observability
is
(4-385)
ytff)
conditions,
initial
we obtain
the transfer
Jl. U(s)
YJis)
S
(4-383)
C(s)
Now
if it is
(4-386)
A,-
is
FY(5)
DPY(i)
(4-387)
assumed that
D=
d2
[rf 1
...
(4-388)
d]
then
= DP =
[/
f2
...
/J
(4-389)
djrr
(4-390)
where
/,
for
</,
d 2 X,
Gto
[/i
[/.
written as
/JY(j)
U(s)
f,7,
is
X,
U(s)
(4-391)
158
State-Variable Characterization of
Dynamic Systems
Chap. 4
U(s)
which
- A,)(j -
(s
expanded by
is
of the form
is
A2)
A)
v$>
&^
(4 - 393)
we
(4-394)
(r,=/,y,
Therefore,
when
a,
be zero if/, =
0, y, will
0,
and the
state y, is uncontrol-
lable.
For
observability,
was established
it
earlier that
from Eq.
for
F must
i
2,
n.
However,
(4-394),
f =
t
(4-395)
f
t
When
a,
4.18
0.
if y,
a,
^ 0.
When
where
f [x(r),
r(f)]
denotes an n
(4-396)
p x
is
As a simple
linear
illustrative
= x,(/) +
x\{t)
j
= x,(0 + r{t)
*$>
(4 - 397)
Sec. 4.18
it
159
would
now
described. All the terms of the Taylor series of order higher than
and
discarded,
linear
are
all
x (t)=f (x
i
,i
(7/Xx, r)
53+ j=x
S
OXj
1,
2,
n.
x 0J )
0y
r 0J )
(4-398)
*
i
(Xj
Sfjjx, r)
Let
Axi
X,
Ar,
r,
Oj
(4-399)
r 0/
(4-400)
X 0!
(4-401)
r o)
(4-402)
and
Then
Ax,
x oi
= /i(x
x,
Since
Equation (4-398)
is
~~
'
The
last
equation
o>
written
dMx,r)
dx,
j=x
may be
j"
io,r,
or
An
(4-403)
Ax
= A* Ax +
B*Ar
(4-404)
where
'df
dx
dx 2
df{
dx n
dx
dx 2
dx
9fx
dh
A*
B*
El
dL
df
dx
dx 2
dx
dfx
dr,
dli
dr z
df2
dr t
d_h
dr 2
dh
dL
dr,
dr 2
(4-405)
df~
"
drp
df2
'
dr
d_L
dr a
(4-406)
160
/ State-Variable Characterization of
Chap. 4
Dynamic Systems
where
it
just
described.
Example
4-31
Xi
*i=fi=xi
X 2 = f2 = u
(4-407)
(4-408)
1
It
x = Ax + Bu
*l
Fig. 4-27.
(1
e-*i*'i)
SGN
is
represented by
jf,
(4-409)
where
+1
SGN xi =
[
-1
jc,
>
<0
Substituting Eq. (4-409) into Eq. (4-408) and using Eq. (4-403),
(4-410)
we have
the
Aii
Ax 2
At
-M-Ax 2 = Ax 2
(4-411)
^Ax
(4-412)
ax 2
ax
= Ke-V'^Axi
where x 01 denotes a nominal value of x Notice that the last two equations are linear
and are valid only for small signals. In vector-matrix form, ihese linearized state
x
~Ax{
Ax 2
"Axi"
(4-413)
Ax 2
where
a
Ke- K \*"\
constant
(4-414)
Ax 2 =KAxi
Thus the
linearized
model
is
(4-415)
Sec. 4.19
gain K.
On
if
is
on the saturated portion of the nonlinearity, and a = 0. This means that any small
variation in x, (small Ax,) will give rise to practically no change in Ax
2
lie
Example 4-32
In the
last
invari-
= ^t
(4-416)
x2
(4-417)
x,
We
would like to linearize these equations about the nominal trajectory [x ,(0,
X02O)], which is the solution of the equations with the initial conditions x,(0) = x 2 (0)
= 1 and the input u{t) = 0.
Integrating both sides of Eq. (4-41 7),
Xi
Then Eq.
we have
x 2 (0)
(4-418)
(4-416) gives
xi
= -/ +
(4-419)
Therefore, the nominal trajectory about which Eqs. (4-416) and (4-417) are to be
linearized is described by
*oi(0
x Q i(t)
Now
=
=
~t
1 =
V
dx
(4-420)
(4-421)
dxi
we
get
du
<?x.
Xi
Ax,
= -3-Ax
X02
Ax 2 =
Substituting Eqs. (4-420)
(4-422)
oAx,
,Ai/
(4-423)
which
4.19
is
Ax,
"0
_Ax 2
_o
2"
o_
"
Ax,
_Ax 2 _
"
J -'_
(4-424)
when
is
by means of discrete
state equations.
situations.
The
first
one
is
that the
data elements, but the signals at certain points of the system are discrete or
discontinuous with respect to time, because of the sample-and-hold operations.
In this case the components of the system are
still
described by differential
may be
162
Chap. 4
Dynamic Systems
generated from the original differential equations. The second situation involves
systems that are completely discrete with respect to time in the sense that they
discrete data only, such as in the case of a digital controller
Under this condition, the system dynamics should be
computer.
or digital
receive
Let us consider the open-loop discrete-data control system with a sampleand-hold device, as shown in Fig. 4-28. Typical signals that appear at various
points in the system are also shown in the figure. The output signal, c(t),
rit)
r*(t)
'
J\-J-
(t)
Zero-order
hold
h(t)
c(t)
i:
T IT 3T AT 5T6T IT
-*-
<TTTv
*-
^liJ-
*-
::
Sec. 4.19
ordinarily
is
h(kT)
r(kT)
0,
2,
163
h{t),
(4-425)
Now we let the linear process G be described by the state equation and output equation
ML = Ax(0 + Bh(t)
c(t)
where
earlier.
x(t)
>
+ EA(0
Dx(0
signals, respectively.
for
tft
and output
which
)\(t
(4-427)
("
<f>(t
x)Bh(x) dx
of the
(4-428)
we
If
Using Eq.
(4-426)
we
let
(k
l)rand
kT. Then
where
l)r]
sign.
4>(T)x(kT)
+ |^
VT
+l)T-
$[{k
t]BA(t)
(4-429)
d-c
Since h(t)
(k
is
h{kT)
is,
Equation (4-429)
is
= r(kT)
for
kT<t<
written
+Ur
x[(*
\)T]
${T)x{kT)
+ |2
x[(fc
l)r]
tt7>(*r)
*K*
l)r
- t]B dx r(kT)
(4-430)
or
Q(T)r(kT)
(4-431)
where
9(T)
=
J kT
Equation (4-431)
form. Since
is
^k +
1}
T~
T]B dZ
it is
referred
k
*
k
=
=
=
= (K7>(0) + 9(7X0)
x(2r) = <K7>(T) + e(7XT)
x(3J) = (f>(T)x(2T) + 0(7X27)
x(T)
0:
1
2:
k= k-
1:
x(kT)
= (K7>p -
1)T]
*(T)r[(k
(4-433)
(4-434)
(4-435)
\)T]
(4-436)
164
State-Variable Characterization of
Chap. 4
Dynamic Systems
Substituting Eq. (4-433) into Eq. (4-434), and then Eq. (4-434) into Eq. (4-435),
solution for Eq. (4-431):
, and so on, we obtain the following
.
.
x(kT)
Equation (4-437)
is
= <f>*(r>x(0) + S p-'-KTWIW.iT)
(=0
(4-437)
defined as the discrete state transition equation of the discretenote that Eq. (4-437) is analogous to its continu-
It is interesting to
data system.
(4-67). In fact,
(4-67) describes the state of the system of Fig. 4-28 with or without sampling.
The discrete state transition equation of Eq. (4-437) is more restricted in that it
= kT (k =
0, 1, 2,
.),
and only
if
where
is
+ N)T] =
4>
N (T)x(kT)
+ 2
4>-'-\T)B(r)r[(k
+ i)T]
is left
as
(4-438)
an exercise
The output of the system of Fig. 4-28 at the sampling instants is obtained by
substituting t = AT and Eq. (4-437) into Eq. (4-427), yielding
c(kT)
Dx(kT)
D4>*(7>(0)
+ Eh(kT)
(4-439)
*-i
+ D 2 p-'-KTWTWT) +
Kh{kT)
An
method
= kT, we get
f()H-A)T
x[(k + A)T] = ftA7X*r) +
<j>P +
\ kT
= 4>(AT)x(kT) + 9(A7>(A:7)
A<
and
between
and
1,
x(t)
t
r]B dx r(kT)
^^
is
+*(T) = ftkT)
which is proved as follows.
Using the homogeneous solution of the
we have
Let
A)r
pling instants
One
<
= kT and
ta
= #t -
x{kT)
Also, by the recursive procedure
is
that
(4-441)
(4-442)
foM'o)
becomes
= ftfcTXO)
= {k +
with
t
(4-443)
\)T and
kT, k
0, 1,
Sec. 4.20
2,
165
x(kT)
(4-444)
<J>*(T)x(0)
In view of the relation of Eq. (4-441), the discrete state transition equations
x(kT)
x[(A:
(4-445)
'"
(4-446)
x[(k
1)7]
= Ax(kT) +
Br(A:r)
(4-447)
where A, B, D, and
two
+ Er(kT)
Dx(kT)
is
same form
basically of the
situations
is
$(T) and
as Eq. (4-431).
The only
(4-448)
8(7") are
is
and
B matrices
itself represents
an out-
(4-431).
x(kT)
A*x(0)
is
A*-'-'BrOT)
i
signals.
written
(4-449)
where
A*
= AAAA...A
(4-450)
4.20
The
x[(k
1)7"]
= Ax(kT) +
Br(&r)
(4-451)
zx(0+)
= AX(z) + BR(z)
(4-452)
(zl
A)" zx(0-r-)
(zl
A)" 'BR(z)
(4-453)
166
State-Variable Characterization of
The
Dynamic Systems
Chap. 4
x(kT)
g- '[(zl
is
A)" z]x(0)
l
- A)
g-*[(zl
BR(z)]
(4-454)
In order to carry out the inverse z-transform operation of the last equation,
write the z-transform of
g(A")
A*
f] A*z"*
k=
=I+
Az"
+A
from the
last equation,
we
z" 2
(4-455)
Az -1 and
we
as
subtracting the
get
(I-
Az-')S(A*)
(4-456)
g(A k )
(I
A*
Az"
)"
(zl
A)"'z
(4-457)
or
l
-A
[{zl
-1
)
(4-458)
Now we
- A)-'BR(z)] = S
and
A^-'-'BrOT)
(4-459)
x(kT)
k-\
+ 2
=
A*x(0)
A*-'-'Br(ir)
(4-460)
x(0+)
X(z)
When
this
equation
is
A)-BR(z)
(4-461)
we have
C(z)
Thus the
[D(zl
A)" 'B
G(z)
D(zl
A)-'B
E]R(z)
(4-462)
is
(4-463)
The
D[adj(zI-A)]B
|zI-A|
is
+ lzI-A|E
{A _
m)
defined as
(4-465)
constant coefficients:
Sec. 4.21
c[(k
1071
+ n - 2)71 +
+ a _ lC [(k + l)r] + a c(kT)
+ VP + m- 1)7] +
n>m
+ b m . A(k + 1)T] + b m r{kT)
fl,c[(fc
+ n-
l)r]
a 2 c[(k
VP + m)T]
C{z)
R(z)
The
z'
(4-466)
sides
167
written
is
+ Z>,z"-' + ... + b m z + b m
+ a^"" +... + o,. z + a,
,.
v
46? ;
is
z"
Example 4-33
fljz"- 1
a_ ,z
(4-468)
equation
c(k
2)
5c(k
1)
3c(/c)
r(k
1)
(4-469)
2r(/t)
initial
conditions yields
z^C{z)
From
5zC(z)
= zR(z) +
z2
R(z)
The
3C(z)
characteristic equation
is
5z
2R(z)
is
(4-470)
easily written
^ *' l)
z1
The
XiQc)
two
two
5z
= c(k)
= x,{k +
x 2 (k
from which we have the
(4-472)
(4-473)
1)
- r(k)
(4-474)
Xl (k
Xl (k)
system as
(4-475)
3r{k)
(4-476)
(4-477)
-3
The same
4.21
-5_
is
obtained by using zl
|
A| = 0.
When
a discrete-data system
state equations,
is
may
state
168
State-Variable Characterization of
Dynamic Systems
Chap. 4
Some of the
operations
and
Multiplication by a constant:
1.
= ax^kT)
X (z) = aX,(z)
2.
x 2 (kT)
(4-478)
(4-479)
Summing:
x 2 (kT)
X (z)
2
= x (kT) + Xi (kT)
= X (z) + Z,(z)
(4-480)
(4-481)
3.
x 2 (kT)
= Xi [(k + l)T)
X (z) = zX,{z) - zx,(0+)
(4-482)
(4-483)
= z- X {z) +
(4-484)
or
X^z)
The
state
(z)
-OX
o-
(z)
(4-484)
(z)
X2 (z) = aX
The
time
rj.
= 0+
Then Eq.
written
is
z'
X (z) + x, {n)
(4-485)
(z)
X2 (z)
Example 4-34
c(k
initial
can be generalized to
Z,(z)
*,(<)+)
2)
fj.
5c(k
1)
3c(k)
r(k
1)
2r(k)
(4-486)
(z)
One way of
X2 (z) = XQ (z) + X
(z)
system
is
9*i(0+)
x,(k
x 2 (k
X2 (z)o
O^(r)
Using
+
+
1)
1)
= Xl (k) + r(k)
= -3 Xl (k) - 5x
essentially the
same
(z)
_1
is
is
always appear as
will
state diagram.
diagram.
3r(k)
(4-488)
used to relate
2 (k)
and (4-488)
(4-487)
x^k +
1) to Xi(k).
Sec. 4.21
R{z)
169
C{z)
-3
diagram of the system described by the difference
equation of Eq. (4-486) or by the state equations of Eqs. (4-487) and
Fig. 4-30. Discrete state
(4-488).
As an alternative, the state diagram can also be drawn directly from the difference equation by means of the decomposition schemes. The decomposition of a discrete
transfer function will be discussed in the following section, after we have demonstrated some of the practical applications of the discrete state diagram.
state transition equation of the system can be obtained directly from the
{z) and
diagram using the gain formula. Referring to
z (z) as the output nodes
and to xi(0+), * 2 (0+), and R(z) as input nodes in Fig. 4-30, the state transition
equations are written in the following vector-matrix form:
The
state
"1
~ A
where
+5z-'
-3Z-
"*i(o+r
1
A=
rz-i(l
5z-'
+5z-')-3z- 2
_ 3z -i_3 z -2
A|_
_* 2 (0+)_
3z-
R{z)
(4-489)
(4-490)
The same transfer function between R(z) and C(z) as in Eq. (4-471) can be obtained
directly from the state diagram by applying the gain formula between these two nodes.
Decomposition of Discrete Transfer Functions
The
is
state
diagrams are
<2f)
(4.491)
Equation (4-491) is used for direct decomposition after the numerator and the
denominator are both multiplied by z~ 2 For cascade decomposition, the trans.
x 2 (0+)
Riz)
C(z)
-3
decomposition
(a) Direct
x, (0 +)
R(z)
C(z)
(b)
Cascade decomposition
*i(0+)
RU)
(c) Parallel
5z
tion, (b)
170
decomposition
(z
2)/(z 2
by the three methods of decomposition, (a) Direct decomposiCascade decomposition, (c) Parallel decomposition.
3)
Sec. 4.22
form
C{z)
R(z)
For the
(4-492)
+ 4.3)(z + 0.7)
(z
C(z)
R(z)
4.22
as
by
form
0.64
0.36
~ z + 4.3 +
(4-493)
0.7
When
is
desired for
all
is
times.
first
in the Laplace
domain
is
written
H(s)
In the time domain, the relation
kT<t<(k +
this
e(kT+)
(4-495)
we need
l)T.
e(kT+). For
(4-494)
-E*(s)
simply
is
h(t)
for
-e-*-
e(kT+)
for
diagram rep-
kT < t <
(k
H(s)
= e(fcr+)
\)T.
The
state
is
(4-496)
diagram representa-
shown
an
hold.
sampled-data system
illustrative
is
constructed,
We
in Fig. 4-32.
state
let
shall
As
diagram of a
us consider the
demonstrate the
=
s
Fig. 4-33.
Sampled-data system.
is
written
(4-497)
172
Dynamic Systems
Chap. 4
(4-498)
e(t)
state
technique. Figure 4-34 illustrates the discrete state diagram of he system through
t
x, (0 +
e(kT+)
Fig. 4-33.
from
diagram
this state
*,[(*
=
c(kT) =
1)7]
e- T x,(kT)
(1
e- T)e{kT)
(4-499)
x (kT)
(4-500)
Therefore, the output response of the system can also be obtained by solving
the difference equation of Eq. (4-499).
If the
c(t) is
desired for
all
t,
x (kT+)
{
,'
e(kT+)
*>
His)
kT<t<(k +
1)T.
interval
Sec. 4.23
To
determine
which
c(t),
is
also *i(f)>
we must
first
kT <
< (& +
have
*(*>
for
We
(4 " 501)
l)r.
173
last
equa-
tion gives
*i(0
kT< < (k +
t
[1
'- kT>
e-
e- -*"jc,(Jtr)
(4-502)
4.23
]e(kT+)
It is
becomes Eq.
if
we
t is
valid for
information on
let t
{k
Y)T
(4-499).
When
it
A(t)x(t)
B(/)r
(4-503)
c(?)
D(0x(r)
E(r)r(?)
(4-504)
where
x(r)
r (0
c (0
=n
=P
=9
state vector
input vector
output vector
generally
^
^
first
a(t)x(t)
(4-505)
a{t)dt
(5-506)
sides to get
In x(t)
In x(t Q )
a(z)
J'
dx
(4-507)
Therefore,
x{t)
where
denotes the
initial
exp
a{x) dx\x(t a )
(4-508)
time.
174
State-Variable Characterization of
Chap. 4
Dynamic Systems
matrix
is
<f>(t, t 9 )
exp
(4-509)
a(x) dx
Notice that for the time-varying case, the state transition matrix depends upon
t
state equation
(t)
it is
A(0x(r)
(4-510)
x(r)
where $(t, t
the problem
) is
is
)x(t
ftr,
satisfies
The question
in general.
(4-511)
is: Is {t,
related
To answer
into a
power
exp T
H"
exp
let
(4-512)
A(t) dx
series,
A(t) dx)
=I+
+^
A(t) dx
[ A(t) dx f
A(<r) </*
...
(4-513)
*-[T.
A(0
A(t) dx
m L'
do
A(t)
j-
A(t)
rft
A(f)
(4-514)
A(0 exp
["
('
By comparison of Eqs.
^ exp
A(t) dx
(4-514)
and
A(0
A(0
we
(4-515),
A(t) dx
we have
j"
A(t) dx
(4-515)
see that
A(0 exp
A(0<K?,
A(t) Jt
(4-516)
or
${t,
dt
if
and only
* o)
(4-517)
A(t) dx A(t)
(4-518)
if
A(r)
f A(t) dx
that
/ )
to
'
Jf la
is,
A(0
and
A(t)
rfr
to
commute.
The requirement
that A(r)
and
its
integral
commute
(4-512) will
is
evidently a very
not be valid.
Sec. 4.23
Most of
4>(t
t ),
1.
2.
3.
<KMo)
1
<f.-
(?,?
I-
listed as follows
,0.
<f.(/
Solution of the
${t 2
for any
t )
t2
moment, we
for the
175
t )
${t, to)f[{t)
(4-519)
where
r\(t) is
satisfies
4>U,
toMO + W,
t )r\(t)
t,
and
simplifying,
we
(4-520)
get
*oM) = Bu(0
(4-521)
Thus
f|(?)
B(f)u(0
<f"'(f, t )
(4-522)
and
0 =
The vector
r\(t ) is
f *
_i
?o)B(t)u(t) dx
= W,
)x(t
(4-523)
tl(f)
(t,
Thus
substitut-
we have
$(r,
$-(*,
)B(t)u(t)
rft
(4-524)
Since
W,
)*-'(T,
r )
= W, /)<K'o, t) =
<K', t)
(4-525)
dt
(4-526)
which
is
ftf,
)x(r
f'
$(t, t)B(t)u(t)
many
t )
is
system
is
One method of
by
discretizing the
176
i(t)
Chap. 4
Dynamic Systems
~ -L{x[(k +
l)T]
kT<t<(k+
x(kT)}
l)T
(4-503)
= A*{kT)x(kT) + B*(kT)r(kT)
over the time interval, kT < < (k + l)T, where
A.*(kT) = TXQcT) + I
x[(k
l)T]
(4-527)
is
approxi-
(4-528)
B*(kT)
TB(kT)
much
the same
way
as in the time-
REFERENCES
State Variables and State Equations
1.
2.
B. C.
New
York, 1967.
3.
D. W. Wiberg, Theory and Problems of State Space and Linear Systems (Schaum's
Outline Series), McGraw-Hill Book Company, New York, 1971.
R. B. Kirchner,
"An
Explicit
Formula
W. Everling, "On
p. 413,
6.
7.
the Evaluation of e A by
10.
Series," Proc.
T.
M.
55,
Vidyasagar,
by Truncated Power
Amer.
of Evaluating eA!
in
Closed Form,"
IEEE
A
C. G. Cullen, "Remarks on Computing e ',"
Vol. AC-16, pp. 94-95, Feb. 1971.
J.
IEEE, Vol.
Power
Mar. 1967.
'
IEEE
Chap. 4
References
11.
177
Vol. 120,
Transformations
C. D. Johnson and
12.
ical
IEEE
the Transformation to
Canon-
13.
I.
14.
IEEE Trans.
L.
15.
M. Silverman, "Transformation
IEEE
17.
18.
S. J.
tems,"
19.
20.
21.
wood
Inc., Engle-
1967.
Diagram
State
22.
Cliffs, N.J.,
a Matrix,"
B. C.
Kuo,
Systems,"
Controllability
"State Transition
WESCON Convention
and Observability
23.
24.
IRE
Trans. Automatic
189-213, 1962.
25.
26.
27.
/.
178
28.
E.
Kreindler and
ability
Chap. 4
Dynamic Systems
P.
Sarachik, "On the Concept of Controllability and ObservIEEE Trans. Automatic Control, Vol. AC-9, pp.
of Linear Systems,"
A. R. Stubberud,
IEEE
30.
31.
"A
Trans. Application
and
Nov. 1964.
IEEE Trans.
"An
White
Plains,
N.Y.
PROBLEMS
4.1.
electric
networks shown
in Fig. P4-1
4.2.
The following
e(t)
(state
Chap. 4
Problems
(a)
*!)
dt 1
(b)
rf
(c)
(d)
5c(t)
c(Q
rfc(/)
c{t)
6^ +
+1_u
"
Using Eq.
r(t)
r(t)
dr(t)
c(T)dT
5c(0
rfr
(4-42),
/(/)
,(/)
<fr
show
2^
= ,H0 +
that
Ar
-r
Ax(0
-1
<((/)
"1"
B =
-2_
_1_
"0"
(b)
-2
-3_
-2
0"
B=
_1_
B =
-2_
-1
(d)
+ B(0
0"
(a)
(c)
J,,
iO)
Klf2
.
T\
ji^
Find the
"10"
_ 1_
0"
-1
ro
B=
-1
4.5.
Find the
for
4.6.
> 0.
Given the
is
is
Problem 4.4
state equation
= Ax(0 +
(t)
B(0
where
~0
A=
r
-1
_2
0_
y(0
where A! and
4.7.
For the
A,x(/)
becomes
B!(0
Problem
4.6, if
B=
4.8.
179
f.
((>(0
4.4.
-c(t)
"+
^
rf/
4.3.
c(t)
rfM)
<ft
(e)
+ 3*0
can the state equation be transformed into the phase- variable form? Explain.
Given the state equations of a linear time-invariant system as
x(/)
= Ax(?) +
B(0
180
State-Variable Characterization of
Dynamic Systems
Chap. 4
where
0~
ro
B=
-2
-3.
determine the transfer function relation between X(s) and U(s). Find the
eigenvalues of A.
4.9.
show
that
1
- A)B =
adj (si
and the
characteristic equation
s"
4.10.
+ ais"
A closed-loop control
-1
is
+ a 2 s" -2 +
system
is
+ aa -is +
a == 0.
described by
= Ax(0 + Bu(0
u(/) = -Gx(0
u(/) = /--vector, A is n
(t)
where
r
x(t)
= //-vector,
matrix.
Show
A BG.
Let
x n feedback
eigenvalues of
n,
is
r,
and
is
the
0"
1
1
A=
B=
1
-2
G=
[i
g2
-5
gi
-10_
g*\
+/1. Can
all
the eigenvalues of
A BG
problem?
4.11.
tion
(a)
is
Find the
d 2 c(t)
Ml)
dt 2
dt
1,
t(0)
= 0,
and
c(t)
= r(f)
<J>(/).
r(t)
4.12.
equations
system
is
Chap. 4
Problems / 181
rf'cKQ
rf/
+
,
dc,(rt
+ 2c,(0-2c 2 (0=r,(0
rf/
4.13.
c 2 (0
r 2 (f)
(a)
Write the state equations of the system in vector-matrix form. Write the
output equation in vector-matrix form.
(b)
Find the transfer function between the outputs and the inputs of the system.
Given the
state transition
a and
co are real
jo
(c)
Given the
numbers.
(a)
(b)
= Ax(/), where
equation (t)
A=
4.14.
ci(0
dt 2
<^(/).
a linear system as
state equations of
= Ax(/) + B(0
(t)
where
ro
on
B=
-11
-6
x(t)
4.15.
Given a
-6J
1,X 2
|_1.
A into
-2, X 3
3. Find a transformation
A = diag [Xi X 2 X 3 ],
a diagonal matrix
1(0
= Ax(f) +
by
Bm(0
where
'
0'
B=
.-25
The eigenvalues are X
-35
_1_
11_
I, X 2
ri
5, X 3
= Py(/) so that A is
transformed state equations are
x(/)
HO = Ay(0 + T(0
Find
4.16
Draw
(a)
and T.
state
(0
= Ax(/) +
Bu(t)
-3
A = -1
-5
(b) (r)
Ax(f)
0"
-1
-2
-1
ro
B=
l
4.17.
P4-17.
182
Dynamic Systems
State-Variable Characterization of
Chap. 4
Figure P4-17.
(b)
(c)
Draw
4.18.
Draw
state
decomposition
(a)
<
b>
G(s)
10
s3
5s 2
6(s
+ 4s +
10
1)
W = s(s + 1F+
3)
Write the state equations from the state diagrams and express them in the
phase-variable canonical form.
4.19.
Draw
state
position
+ 1)
+ 2)(s +
6(s
(a)
G(s)
s(s
<i3P + ^>
3)
5c(0
-f- +
Write the state equations from the state diagrams and show that the states are
decoupled from each other.
4.20.
Draw
state
decomposition.
4.21.
linear system,
iQfr + i)
G(s) =
UW
(s + 2(s + 5)
Draw
state
position.
4.22.
The
state
diagrams for the system using three different methods of decomstate diagrams should contain a minimum number of integrators.
The
is
shown
Figure P4-22.
in Fig. P4-22.
Problems / 183
Chap. 4
(a)
Assign the state variables and write the dynamic equations of the system.
(b)
4.23.
Draw
4.24.
The
state
state
network shown
electric
shown
is
in Fig. P4-1.
in Fig. P4-24.
Figure P4-24.
if
4.25.
Given the
state equation
(0
= Ax(?)
where
-2
0"
-2
.0
(a)
(b)
4.26.
-2_
Given the
<J)(0-
state equation
(/)
Ax(t)
B(0
where
ro
on
B =
-4
-2
-3J
|_1.
"-1
0"
-1
-1
4.27.
Given the
linear system
(0
where
(/) is
= Ax(0 +
B(0
The
=P"iAP
-1_
state transition
<J>(/)
= -Gx(0
e (A-BG)
-![(,,!
e (A-BG)t
gAtg-BGt
_ a + BG)-
is
1
184
Dynamic Systems
Chap. 4
where
e At
e -BGr
= -l[(5I _ A)"']
= -i[(sl + BG)"
Determine the
system shown
in Fig. P4-28.
Figure P4-28.
(a)
(b)
= 1, b = 2, c = 2, and d=\.
Are there any nonzero values for
a
a, b, c,
is
not
Figure P4-29 shows the block diagram of a feedback control system. Determine
and
by the following
It
\X
+ 2
'
'
x2
.9
(a)
matrices.
4.30.
The
Coupling of
states.
C(s)
R(s)~
(a)
s
s3
is
given by
+a
+6s 2 + Us +
is
either uncontrollable or
unobservable.
(b) Define the state variables so that
(c)
4.31.
one of them
is
uncontrollable.
= Ax(0 + B(0
is
unobservable.
Problems
Chap. 4
where
o
L-l
a_
"
185
"1"
B =
L*J
such that the system
is
completely
controllable.
4.32.
Draw the state diagram of a second-order system that is neither controllable nor
observable.
4.33.
b\,
b2
d u and d2
is
(0
= Ax(0 +
~i
r
_o
c(0
D=
4.34.
Bk(/)
i_
Dx(0
d2 ]
[/,
The block diagram of a simplified control system for the Large Space Telescope
(LST) is shown in Fig. P4-34. For simulation and control purposes, it would
be desirable to represent the system by state equations and a state diagram.
Control moment
gyro dynamics
Gimbai
controller
Vehicle
K,
Command
Kp s +
~\+ K
-J
K,
+ Kj
position
Js 2
JG s
1
Vehicle
dynamics
KN
Figure P4-34.
(a)
Draw a state diagram for the system and write the state equations in
vector-
matrix form.
(b)
(c)
- gjXi g 2 x 2 g]X - g x Xi
3
Find the values of gi, g 2 gi, and g^ such that the eigenvalues of the overall
system are at s = 100, 200, 1 +/1, 1 jl. The system parameters
= 600, Kj = 9700, JG = 2, Jv = 10 5 Kp = 216, and
are given as
,
KN = 300.
186
State-Variable Characterization of
4.35.
The
Dynamic Systems
Chap. 4
difference equation of
c[(k
+ 0.5c[(A: +
2)T]
1)T]
(a)
(b)
The
k =
2, 3,
0.1c(kT)
given by
initial
.
is
10 by
+
+
Xl (k
x 2 (k
state equations,
1)
1)
= 0.1x 2 (k)
= -x^k) + 0Jx
(a)
(b)
>it)
+ r(k)
</>(&).
C(z)
R(z)
4.38.
2 (k)
(z
l)(z 2
-z
3)
Draw
tO ^T
- e*(0
e(t)
in vector-matrix form..
z.o.h.
is
2Q +
s(s
shown
c(t)
0.5)
+ 0.2)
Figure P4-38.
Draw
(b)
(c)
(a)
x[(fc
1)71
T=
= <J>(r)x(/cD + B(T)r(kT)
0.1 sec.
in Fig. P4-38.
*-
5
Mathematical Modeling
of Physical
5.1
Systems
Introduction
One
of the most important tasks in the analysis and design of control systems
is
some
extent.
if
linearity.
how
important to point out that the modern control engineer should place
on the mathematical modeling of the system so that the analysis
and design problems can be adaptable for computer solutions. Therefore, the
main objectives of this chapter are
It is
special emphasis
1.
2.
188
Chap. 5
encountered in practice.
5.2
The
classical
way of
is
the
loop method and the node method, which are formulated from the two laws of
Kirchhoff. However, although the loop and node equations are easy to write,
they are not natural for computer solutions.
network equations
is
equations of electrical
in this section.
networks
RLC network
of Fig. 5-1 to
V\AA<
T1KP
of
HO
+
e(t)
e c (t)
-J?
RLC network.
Fig. 5-1.
It is relatively
where
q(t)
is
= L d-^l + R d
-f + -L q
is
q{t)=\'
It is
shown
in
(5-1)
(t)
i(t)
by
(5-2)
i{-c)dx
differential
equation in
Eq. (5-1) can be replaced by two first-order differential equations called the
state equations. In this case it is convenient to define the state variables as
Xl (t)
where ec (t)
is
= ^P =
* 2 (0
/(0
(5-3)
e c (t)
and
= C^)
(5-4)
= L **M + Rx
2 (t)
Xl (t)
(5-5)
Sec. 5.2
Thus, from Eqs. (5-4) and (5-5) the state equations of the network are
dx
(t)
*,<0
dt
^r -
(5-6)
Rx
x,(t)
e(t)
2 (t)
(5-7)
A more direct way of arriving at the state equations is to assign the current
in the inductor L,
variables.
Then
i(t),
C and
the voltage across L in terms of the state variables and the input source. This
way
the state equations are written by inspection from the network. Therefore,
=
C^4^
dt
Current in C:
Since x^t)
Voltage across
and x 2 (t)
e c (t)
In general,
it is
L di(f)_
i(i), it is
and
(5-8)
i(t)
-e c (t)
Ri(t)
e(t)
(5-9)
(5-7).
One must
recognize that the basic laws used in writing state equations for
networks are
electric
method using
Example
As another example of writing the state equations of an electric network, consider the network shown in Fig. 5-2. According to the
5-1
is
available. 1
foregoing discussion, the voltage across the capacitor e c and the cur/, and i2 are assigned as state variables, as shown in Fig. 5-2.
Fig. 5-2.
Network
in
Example
5-1.
The state equations of the network are obtained by writing the voltages across the
inductors and the currents in the capacitor in terms of the three state variables. The
state equations are
L2
dh(t)
dt
-Rihit)
+ e (t)
c
e(.t)
(5-10)
(5-11)
190
Chap. 5
c^ = m-m
Rearranging the constant
(5-12)
canonical form:
1
_*i
diAtY]
-+-
dt
dhU)
dt
hit)
5.3
e(t)
(5-13)
edt)
de c (t)
dt
hit)
Most feedback
From
ponents.
comand
mechanical elements are analogous. In fact, we can show that given an electrical
device, there is usually an analogous mechanical counterpart, and vice versa.
The analogy, of course, is a mathematical one that is, two systems are analo;
if
Motion
forces
= Ma
(5-14)
is
given by
M
where g
is
tion of free
(5-14)
and
fall.
Three consistent
(5-15)
body due
sets
Sec. 5.3
Mass
Units
Weight
Acceleration
Force
m/sec 2
MKS
newtons/m/sec 2
CGS
dynes/cm/sec 2
newton
dyne
cm/sec 2
newton
dyne
British
lb/ft/sec 2 (slug)
lb
ft/sec 2
lb
y(t)
acting
fit)
Ma(t)
= M<?M
M-dv(t)
r
is
where
the acceleration,
is
(5-16)
dt
Force-mass system.
a{t)
is
written
fit)
Fig. 5-3.
191
all
and
applied force.
2.
Linear spring
an element that
some
all
extent.
f{t)
where
is
considered
is
analogous to a
is
small,
is
Ky{t)
its
(5-17)
The
three unit
MKS
CGS
newtons/m
dynes/cm
British
lb/ft
is
directly
spring.
y{t)
K
-
is
AAAAA/v-
fit)
If the spring
is
f{t)-T=Ry{t)
Whenever
there
is
(5-18)
motion or tendency
The
frictional forces
encountered in physical systems are usually of a nonlinear nature. The characteristics of the frictional forces between two contacting surfaces often depend
on such factors as the composition of the surfaces, the pressure between the
surfaces, their relative velocity, and others, so that an exact mathematical
description of the frictional force is difficult. However, for practical purposes,
frictional forces
friction,
192
Chap. 5
static friction,
and Coulomb
friction.
in
the following.
1.
y(t)
Dashpot
fit)
f(f)
where
sions of
friction
often repre-
The
is
B&
(5-19)
is
is
H
Fig. 5-5.
by a dashpot such
is
The
The dimen-
Units
MKS
newton/m/sec
CGS
dyne/cm/sec
British
lb/ft/sec
and
frictional force
2.
velocity.
to prevent
can be
where (^)^,
when
the
is
body
of the friction
(F,),_
(5-20)
is
The
direction of velocity.
The
f
+ F.
Slope =
(c)
(b)
is
static
sign
initial
Viscous
Coulomb
friction.
forces,
Sec. 5.3
193
/w-'-GF/ISD
where
is
Coulomb
the
friction coefficient.
is
Rotational Motion
axis.
and
density.
its
For
metric axis
is
J
where
Example
the
is
5-2
its
inertia
= \Mr
shaft
is 1 in.
Wr 2
it
and
is
r is its radius.
is
(5oz)(l
and weighing 5
in) 2
386 in/sec 2
(5-23)
0.00647 oz-in-sec 2
given in weight per unit volume. Then, for a
inertia is proportional to the fourth
first
length. Therefore,
if
the
expressed as
W = p(nr
where
(5-22)
geo-
is
weight
its
given by
2 h)
(5-24)
is
pn^ = omQ6p h/
(5-25)
J
For aluminum, p
is
written
1.56 oz/in 3
(525)
becomes
= 0.0184 hr 4
(5-26)
= 0.00636 hr*
(5-27)
194
Chap. 5
When
T(-^
as
shown
a torque
is
is
written
no = Mt) = J ^r = J ~$P
<' >
The
tities
Units
(5-28)
Inertia
Torque
Angular Displacement
radian
MKS
kg-m 2
CGS
g-cm 2
newton-m
dyne-cm
English
slug-ft 2
lb-ft
radian
or lb-ft-sec 2
oz-in-sec 2
oz-m
radian
found useful
Angular displacement.
Angular
= 11? =
rad
57.3
velocity.
1
rpm
= ?
= 0.1047 rad/sec
oU
rpm
6 deg/sec
Torque.
=
lb-ft =
oz-in =
g-cm
0.0139 oz-in
192 oz-in
0.00521 lb-ft
Inertia.
1
g-cm 2
lb-ft-sec
oz-in-sec 2
g-cm-sec 2
1
Torsional spring.
lb-ft-sec
As with
1.417
=192
=
=
=
10" 5 oz-in-sec 2
oz-in-sec 2
386 oz-in 2
980 g-cm 2
32.2 lb-ft 2
7X0
-M/WW
of a rod or a shaft
when
it
is
subject to
an applied
torque.
0(0
Fig. 5-8. Torque-torsional spring
system.
tion:
7X0
KB{t)
(5-29)
Sec. 5.3
The dimension
for
195
MKS
CGS
newton-m/rad
dyne-cm/rad
British
oz-in/rad
is
modified to
T(t)
-TP= Kd(t)
(5-30)
Eqs. (5-19),
parts
no _
n dftt)
no
{F )d=0
(5-31)
dt
(5-32)
(5-33)
where
C\)<to
B is
is
Relation
Between
and
Translational
the
Coulomb
friction coefficient.
is
it is
For
momove
instance, a load
may be
controlled to
along a straight line through a rotary motor and screw assembly, such as that
shown in Fig. 5-9. Figure 5-10 shows a similar situation in which a rack and
no,
Motor
HO
-x(0
x(0
/VWWWWWWWWWWWWWVAA
Drive
motor
196
Chap. 5
pinion
is
control
is
as that
shown
all
in Fig. 5-11.
and
5-11
can
/=M/- 2 =
(5-34)
If the radius
motor
using Eq. (5-34) as the equivalent inertia for the system of Fig. 5-9,
2
(5 " 35)
= fte)
= inertia (oz-in-sec
W weight (oz)
L = screw lead (in)
g = gravitational force (386.4 in/sec
2
form of
kinetic
Wk = \Mv
The following
tion:
(5-36)
consistent sets of units are given for the kinetic energy rela-
Sec. 5.3
Energy
Units
Mass
197
Velocity
MKS
joule or
newton/m/sec 2
m/sec
CGS
newton-m
dyne-cm
dyne-cm-sec 2
cm/sec
British
ft-lb
lb/ft/sec 2
ft/sec
(slug)
W = ya>
is
written
(5-37)
where /
the
is
moment
of intertia and
co the
Energy
Units
Angular Velocity
Inertia
MKS
joule or
CGS
newton-m
dyne-cm
gm-cm 2
rad/sec
British
oz-in
oz-in-sec 2
rad/sec
kg-m 2
rad/sec
work required
by y
is
W = \Ky*
(5-38)
where
stored
is
Wp = \KQ*
(5-39)
When
dealing with a frictional element, the form of energy differs from the
previous two cases in that the energy represents a loss or dissipation by the sys-
tem in overcoming the frictional force. Power is the time rate of doing work.
Therefore, the power dissipated in a frictional element is the product of force
and
velocity; that
is,
P=fi>
Since
/=
MKS
unit for
Bv, where
P
The
power
is
(5-40)
is
Bv*
(5-41)
horsepower
(hp).
CGS
system
it is
represented in ft-lb/sec or
Furthermore,
1
Since power
is
is
= 746 watt
= 550 ft-lb/sec
hp
is
(5-42)
is
W = BJv
d
dt
(5-43)
198
Gear
Trains, Levers,
Chap. 5
is
a mechanical device
0,
>
r,N z
2.
The distance
=r N
1
is
proportional to the
(5-44)
is
the same.
Therefore,
6',!,
3.
(5-45)
62 r2
=T
62
(5-46)
JV,
Ti
Ft
N,
r,,,
T, 0,
M^4
4
Tiy
<C2 "2
d7
JV,
Fig. 5-12.
JV,
Gear
Fig. 5-13.
train.
two
gears, co 1
and
co 2 , are
T2
(5-47)
0t
#2
CO!
r2
In practice, real gears do have inertia and friction between the coupled gear
which often cannot be neglected. An equivalent representation of a gear
teeth
elements
is
shown
in Fig. 5-13.
Sec. 5.3
T=
6 1, 6 2
Ti,T2
/, J2
,
=
=
199
applied torque
angular displacements
torque transmitted to gears
= inertia of gears
N u N number of teeth
F = Coulomb friction coefficients
B u B = viscous frictional coefficients
2
F,i,
c2
T2 (t) = J2
The torque equation on the
r(0
By
= /,
^ ^
for gear 2
is
+B
side of gear
+ Fc2 -^
(5-48)
1 is
(5-49)
ei
(a
^- + *, *M> + F M- + 7\(?)
is
w
T
written
converted to
W ST + n F
ddti)
if
Nl
>
<>
e2
(s 50)
(5
|^y
Equation (5-50) indicates that it is possible to reflect inertia, friction, (and compliance) torque, speed, and displacement from one side of a gear train to the
other.
2 to gear
when
reflecting
from gear
Inertia:
(i)V2
Torque:
^T
B2
(--^1
Angular displacement:
Angular velocity:
-jrr-d 2
~co 2
Jy 2
Coulomb
frictional torque:
-rrFc2
c 2
.
,
|a>2l
(NJN2 )
in reflecting
Eq. (5-49),
we
get
HO = /,.
^0 + B U ^M + T
(5-51)
where
= Ji +
J
(jff
Bu = B +
B
Ju
(5-52)
(jfy
(5-53)
200
Example
Chap. 5
5-3
inertia
through a
reflected inertia
Coulomb
on
friction is (^)2
Timing
= 0.4
is
Coulomb
fric-
on the load
oz-in-sec 2
side).
The
The
reflected
oz-in.
belts
(Ni/N2 = ^ with
x 0.05 = 0.002
()
5 gear train
and
serve the
same purposes
except that they allow the transfer of energy over a longer distance without using
an excessive number of gears. Figure 5-14 shows the diagram of a belt or chain
drive between two pulleys. Assuming that there is no slippage between the belt
and the pulleys, it is easy to see that Eq. (5-47) still applies to this case. In fact,
the reflection or transmittance of torque, inertia, friction, etc.,
is
similar to that of
a gear train.
*~/i
T2
02
The
lever system
force in the
(5-55)
x{t)
\
~*~ 2
.HO
In addition,
Output
ical
it
model of backlash
Sec. 5.3
is
201
reference position.
upon the
the output
member
motion
member
When
is
will
the output
stand
still
is
is
is
member
reverses
its
member
will
which time
it is
up on the other
side, at
assumed that the output member instantaneously takes on the velocity of the
input member. The transfer characteristic between the input and the output
displacements of a backlash element with negligible output inertia
Figure 5-17.
To
is
shown
in
members,
let
is
The displacements and velocities of the input and output members are illustrated as shown in Fig. 5-18. Note that the reference position of the
two members is taken to be that of Fig. 5-16, that is, with the input member
respect to time.
when motion
member
and y(0)
is
For
Fig. 5-18,
it is
assumed that
member on
is
so small that
202
Chap. 5
without
friction.
ber. At that time the output member will again assume the velocity of the input
member. The transfer characteristic between the input and the output displacement of a backlash element with negligible output friction is shown in Fig.
5-19. The displacement, velocity, and acceleration waveforms of the input and
output members, when the input displacement is driven sinusoidally, is shown in
Fig. 5-20.
and
5-20.
lie
between those of
Sec. 5.4
203
Displacement
Velocity
Acceleration
Output
Input
is
5.4
driven sinusoidally.
5-4
Let us consider the mechanical system shown in Fig. 5-2 1(a). The
free-body diagram of the system is shown in Fig. 5-21 (b). The force
d 2 y(t)
y)
*-/(/)
(a)
(b)
Free-body diagram.
204
Chap.
written
is
(5-56)
This second-order differential equation can be decomposed into two first-order state
equations, using the method discussed in Chapter 4. Let us assign x x = y and jc 2 =
dy/dt as the state variables.
Then Eq.
(5-56)
is
written
dxj(t)
dt
T
It is
electric
not
**(/)
(5-57)
-i^<)-M^> + ]>>
(5-58)
mechanical system
it is
is
analogous to a
series
RLC
from the mechanical system using a different set of state variables. If we consider
mass is analogous to inductance, and the spring constant K is analogous to the
inverse of capacitance, 1/C, it is logical to assign v(t), the velocity, and fk (t), the force
acting on the spring, as state variables, since the former is analogous to the current in
an inductor and the latter is analogous to the voltage across a capacitor.
ly
that
Then
M Mp =
Force on mass
Velocity of spring
first state
equation
dfk (t)
dt
is
_ Mt)
-i-/(0
-A(0
(5-59)
_ v{t)
(5-60)
on the voltage
through a capacitor.
This simple example further illustrates the points made in Chapter 4 regarding
the fact that the state equations and state variables of a dynamic system are not unique.
Example
is
From
y 2 U)
VA
CF
Md
yi(0
y 2 (t)
dt
y 2 U)
yi(t)
-nptfs-
fU)
dy 2 (t)
-nnnnK(y
-y 2
~dT-
(a)
(b)
Fig. 5-22.
5-5. (a)
Mass-spring-friction
/(
Sec. 5.4
fit)
= K\y
- y 2 (t)]
(t)
let
205
(5-61)
K[yi(t)
Now
(5-62)
us write the state equations of the system. Since the differential equation of
the system
is
way
is
to
decompose
this
(t)
=y
2 (t)
dxi(t)
dt
=x
As an
alternative,
we can
= dy 2 (t)/dt,
(5-63)
2 (t)
+ 3^/(0
-*(')
dt
state variable,
and x 2 (t)
(5-64)
dv(i)
,,,
M as one
state variable, so
we have
-,.
(5-65)
and
/*(')
=/(')
(5-66)
two equations
in Eqs. (5-65)
and
(5-66) are
it
The
situation
is
network analogous
ydt)
where
yi(t)
is
v(t)
dx
+ y 2 (0+)
(5-67)
.^(O-l-)
we can
= &jp + y 2 (t) = + P
is
the
initial
Example
5-6
In this example the equations for the mechanical system in Fig. 5-24(a)
are to be written. Then we are to draw state diagrams and derive
transfer functions for the system.
for the
206
Chap. 5
^^^^^
K-,
M-,
y 2 (f)
K,
*i(>-i
M,
Vlit)
no
/u)
(b)
(a)
Fig. 5-24.
5-6.
equations for the system are written directly from the free-body diagram:
(5-68)
(5-69)
dt*
We may now
dt
differential
equations
x,
= yi
dy\
Xl
dt
(5-70)
dx
x 3 =yi
xt
(5-71)
dt
(5-72)
dy 2
dx 3
dt
dt
(5-73)
Equations (5-71) and (5-73) form two state equations naturally; the other two are
obtained by substituting Eqs. (5-70) through (5-73) into Eqs. (5-68) and (5-69), and
rearranging
we have
dxi
dt
= x2
(5-74)
Sec. 5.4
dx 2
Xi)
dt
dx 3
If
we
Xi)
Wm
207
(5-75)
Xi
dt
dt
~ Ml {Xz ~
(5-76)
-M
Xl
A",
X2
AT2
"
*33
(,
A/ 2
and
+ Bi )x i
(5-77)
written
= *i(0
(5-78)
* = *s(0
(5-79)
yi(t)
The state diagram of the system, according to the equations written above, is
drawn as shown in Fig. 5-25. The transfer functions Y (s)IF(s) and Y2 (s)/F(s) are
x
By/Mj
5-24.
obtained from the state diagram by applying Mason's gain formula. The reader
should verify the following results (make sure that all the loops and nontouching loops
are taken into account)
m
x
2 s*
(s)
MO
F(s)
(Bi
+ B )s +
A
2
(K
+ K2
(5-80)
+K
(5-81)
where
A =
MM
x
+B
2 s*
1 (B l
+ [M (B + B2 + B M2 - Bfis* + [M (K + K2 ) + K M2
+ B2 ) - BiK^s 2 + [K B 2 + B (K + K2 )]s + K K2
)
(5-82)
The state equations can also be written directly from the diagram of the mechanical
The state variables are assigned as v = dy /dt, v 2 = dy 2 /dt, and the forces on
the two springs, /x! and fK2 Then, if we write the forces acting on the masses and the
velocities of the springs, as functions of the four state variables and the external force,
system.
Force on
x :
dv
x
~di
-B lVl
+B v -fK1 +f
x
(5-83)
208
Force on
Velocity on
on
Velocity
Chap. 5
K2
= fi,^i
-%*
df
Ki
"^-
(5-84)
fx:
v2 )
#,(*>i
^p = K v
B2 )v 2 +/ki
(2?i
(5-85)
(5-86)
Example
is
5-7
negligible,
but the
stiffness is
K.
<s^^^^
^^^^
(a)
(b)
Example
5-7.
Assume that a torque is applied to the disk as shown; then the torque or moment
equation about the axis of the shaft is written from the free-body diagram of Fig.
5-26(b):
T(t) = J 4%P +
Notice that
equations
= x 2 (t).
this
system
may be
is
B^ +
K0(t)
(5-87)
x-i(t)
6(t)
and dx
{f)\dt
an
exercise.
5.5
Error-Sensing Devices
in
Control Systems 4
'
The
is
The
error signal
is
sum of
several signals
is
defined in
potentiometer or combination of potentiometers, a differential gear, a transformer, a differential amplifier, a synchro, or a similar element. The mathematical
is
Sec. 5.5
Error-Sensing Devices
in
Control Systems
when
compare two
may be
a voltage
is
When
a constant voltage
its
is
209
proportional
fixed terminals,
potentiometer.
applied across
is
and
potentiometer, the voltage across the variable and the reference terminals will
be proportional to the difference between the two shaft positions. The arrange-
ment shown in Fig. 5-27(b) is a one-potentiometer realization of the error-sensshown in Fig. 5-27(a). A more versatile arrangement may be obtained
by using two potentiometers connected in parallel as shown in Fig. 5-27(c). This
ing devices
(a)
u(0
(b)
o_
-e(0-
(c)
o
Load
Gear
train
T
e(f)
dc amplifier
Input
e a (t)
>U
shaft
(a)
Permanent-magnet
dc motor
Fig. 5-28. (a) Direct current control system with potentiometers as error
210
waveforms of
Sec. 5.5
Error-Sensing Devices
in
can be ac or dc, depending upon the types of transducers that follow the
error sensor. If v(t) is a dc voltage, the polarity of the output voltage e(t) determines the relative position of the two shafts. In the case of an ac applied voltage,
v(t)
the phase of
e(t) acts
case the transfer relation of the two error-sensor configurations can be written
e(t)
= K [O (t) s
6 c (t)]
(5-88)
where
=
K =
e(t)
s
and the
total displacement
if
is
V volts
prevail. In this case the voltage applied to the error sensor is sinusoidal.
is
is
usually
much
The
is
referred to as the
v(t)=V sin
Analytically, the output of the error sensor
e(t)
co c t
is
(5-89)
given by
= K,0JMt)
(5-90)
where 0,(0 is the difference between the input displacement and the load displacement, 0/0
r (?)
c (f). For the 0/0 shown in Fig. 5-29(b), e(t) becomes
a suppressed-carrier modulated'signal.
occurs whenever
212
6c (t)
Chap. 5
Load
Gear
train
Two-phase
ac motor
(a)
(b)
(a).
the signal crosses the zero-magnitude axis. This reversal in phase causes the ac
motor
error 6t).
when a
signal
6 e (t)
is
modulated by a
fact that
Error-Sensing Devices
Sec. 5.5
Control Systems
in
0,(0
sin o>X?)
$K
V[cos(co c
co s )t
To
relations, substitut-
co c
(5-91)
213
cos (co c
+ co
(5-92)
s )t]
no longer contains the carrier frequency co c or the signal frequency co s but it does have the two side bands co c + co s and co c co s
Interestingly enough, when the modulated signal is transmitted through the
system, the motor acts as a demodulator, so that the displacement of the load
will be of the same form as the dc signal before modulation. This is clearly seen
from the waveforms of Fig. 5-29(b).
Therefore,
e(t)
should be pointed out that a control system need not contain all-dc or
components. It is quite common to couple a dc component to an ac component through a modulator or an ac device to a dc device through a demodulaIt
all-ac
For instance, the dc amplifier of the system in Fig. 5-28(a) may be replaced
by an ac amplifier that is preceded by a modulator and followed by a demodula-
tor.
tor.
Synchros.
Among
Synchro Transmitter
is
shown
in Fig. 5-30.
single-phase ac voltage
is
applied to the
Stator
Slip
*1
ac voltage
rings
Rotor
R-,
(a)
Fig. 5-30.
(b)
Stator
214
Chap. 5
which
is
=E
(5-93)
sin co c t
neutral n
is
maximum and
is
written
= KE
e s,Sf)
where
S3 n
A"
is
a proportional constant.
sin cot
(5-94)
the terminals 5[
and
are
e Sm (t)
e sm(0
= KE
= KE
cot
(5-95)
(5-96)
e S,Si
=
=
=
- es =
e Ss =
e Sin ~
e Sl
e Stn
e S,n
\.5KE
sin cot
(5-97)
1.5KE,
sin cot
(5-98)
in
the stator.
Consider
mitter
now
is
tion, as
shown
winding
will
in Fig. 5-31.
displacement 6; that
is,
Es = KE
ESin = KE
Rotor position
of a synchro transmitter.
The magnitudes of
cos (6
cos 9
cos (0
*J~TKE
,s<
^fJKE
sin (9
sin
240)
(5-99)
(5-100)
120)
(5-101)
become
Es
ESl = KE
Fig. 5-31.
The
240)
(5-102)
120)
(5-103)
(5-104)
is
Sec. 5.5
Error-Sensing Devices
in
Control Systems /
215
Volts
Rotor position
6 (degrees)
is
synchro transmitter as
is
is
two
apparently
Synchro
Control
transformer
transmitter
Rotor
R,
Output voltage
proportional to
sin (0,
Fig. 5-33.
dc )
positions, a proportional
generated at the rotor terminals of the control transformer.
Basically, the principle of operation of a synchro control transformer is
voltage
is
is
is
is
cylindrically
essential for a control transformer, since its rotor terminals are usually
electrical device, in
The change
in the rotor
216
Chap. 5
The symbol
CT
is
now
magnetic construction, the flux patterns produced in the two synchros will be
the same if all losses are neglected. For example, if the rotor of the transmitter
is
When
is
in the position
transformer
is
When the
known
in the control
The
shown
in Fig.
shafts of the
two
and
(b).
its
terminal voltage
is
as the
is
at
Control
transformer
Synchro
transmitter
-~~|
/
/f
L/X a = or
!
180
Rotor voltage =
Flux pattern
(b)
Control transformer
Control
transformer
Flux
pattern
Rotor voltage
at
(d)
(c)
is
maximum
among
Sec. 5.5
Error-Sensing Devices
in
Control Systems /
217
it
any position other than that shown in Fig. 5-34(a), the flux patterns will
and the rotor voltage of the control transformer will be pro-
in
shift accordingly,
of the control transformer versus the difference in positions of the rotors of the
transmitter and the control transformer is shown in Fig. 5-35.
Rotor
voltage
Vr
360
Fig. 5-35.
a=(dr -dc
dif-
From
Fig. 5-35
it is
is
a nonlinear
constant
Ks
^e~rri
"
(5 105)
where
E=
=
=
=
Ks =
Br
9C
9e
error voltage
shaft position of synchro transmitter, degrees
shaft position of synchro control transformer, degrees
mitter
connected to the controlled shaft, and the rotor of the synchro transis connected to the reference input shaft. When the controlled shaft is
When an
angular misalignment
exists,
218
Chap. 5
Two-phase
induction
motor
Reference
Controlled
output
input
(a)
or
em
ac
amplifier
+ ^-
Motor
Gear
ec
load
(b)
Fig. 5-36. (a) Alternating current control system
employing synchro
(a).
appears at the amplifier input, and the output of the amplifier will drive the
motor in such a direction as to reduce the error. For small dev iations between the
controlled and the reference shafts, the synchro error detector can be represented
by the constant Ks given by Eq. (5-105). Then the linear operation of the positional control system can be represented by the block diagram of Fig. 5-36(b).
From the characteristic of the error detector shown in Fig. 5-35, it is clear that
Ks has opposite signs at the two null positions. However, in closed-loop systems,
only one of the two null positions is a true null; the other one corresponds to an
unstable operating point.
Suppose
that, in the
Tachometers / 219
Sec. 5.6
motor to rotate
in the
same
still
further
is reached.
In reality, the error signal at the rotor terminals of the synchro control
transformer may be represented as a function of time. If the ac signal applied to
the rotor terminals of the transmitter is denoted by sin a> c t, where co c is known
is
given by
= Ks O (i) sin a
e
e(t) is
(5-106)
signal.
5.6
Tachometers 4
x2
the
'
o
Fig. 5-37.
first state
is
Schematic diagram of a
is
The input-output
tachometer.
e l (t)
where
e,{t) is
=K
(5-107)
^jp
and K,
is
defined as
rpm
or volts
per 1000 rpm. The transfer function of an ac tachometer is obtained by taking the
Laplace transform of Eq. (5-107); thus
g> =
K,s
(5-108)
described above.
220
Chap. 5
excitation voltage
is
and
its
5.7
DC Motors
in
if a
a control system
ac.
Similarly, a
phase-sensitive demodulator
is
by
dc
used
Control Systems
Direct current motors are one of the most widely used prime movers in industry.
The advantages of dc motors are that they are available in a great variety of
types and sizes and that control is relatively simple. The primary disadvantage
of a dc motor relates to its brushes and commutator.
For general purposes, dc motors are classified as series-excited, shunt-
and separately-excited, all of which refer to the way in which the field
is excited. However, the characteristics of the first two types of motor are highly
nonlinear, so for control systems applications, the separately-excited dc motor
excited,
is
ture controlled.
is
ing.
Field-Controlled
DC Motor
armature voltage
Ra
armature resistance
0(f)
air-gap flux
e b (t)
Kb
in Fig. 5-38.
DC
Sec. 5.7
Motors
/.
torque constant
4(f)
armature current
field
current
e f (t)
field
voltage
Tm (t)
Jm
rotor inertia of
Bm
OJt)
To
= constant
K,
i (t)
in
motor
2.
= Kf
0(0
3.
/'
/.
is
is,
(5-109)
i {t)
made:
flux
TJt)
where
Km
is
= Km IJ(t)
(5-H0)
gives
TJt)
If
we
= Km Kf IJf {t)
(5-111)
Km Kf I
(5-112)
let
Kt
<t
TJf)
=K
(5-113)
f {t)
Referring to the motor circuit diagram of Fig. 5-38, the state variables are
assigned as if (t), cojt), and OJt). The first-order differential equations relating
these state variables and the other variables are written
L/-^ = -RMt) + e f
(t)
(5-114)
222
Chap. 5
Jn
dcoJt)
dt
dBJf)
dt
By proper
= _ BmCOm(t) +
TJf)
(5-115)
= co m (t)
(5-116)
in
equations:
'
dif {t)
'
f
i/O
dt
d(Q m (t)
Kj
Bm
dt
~.
(5-117)
efy)
[9Jf)
dt
coJJ)
ddjt)
.
~
1
in Fig. 5-39.
The transfer
"s-l
Ki/Jm
ULf
oe
-R L
f' f
~B m Um
is
obtained from
Ef (s)
K,
Lf Jm s 3
(R,JM
(5-118)
+ Bm Lf )s + R B n s
2
or
Om(s)
Ef(s)
R a Bm s(l +
T m s)(l
(5-H9)
x f 8)
where
r_
xf
= -=
= field electrical time constant of motor
K
-=P-
Armature-Controlled
DC Motor
is
shown
in
it is
relates the
DC
Sec. 5.7
+o
Motors
in
Control Systems
223
WvV
Constant
field
current
armature current,
i a (t)
thus
TJt)
where
K,Ut)
(5-120)
is
emf voltage
is
is
(f>
is
constant also.
Of
The back
=K
e(t)
With reference
dOJf)
dt
(5-121)
co m (t)
Ra
diJLO I
dt
dco m (t)
ao
La
Bm
dBJf)
(5-122)
ej,t)
OJt)
dt
co m (0
dt
The state diagram of the system is drawn as shown in Fig. 5-41 The transfer
function between the motor displacement and the input voltage is obtained from
.
e m (s)
E (s)
a
K,
La Jm s>
(R a Jm
+ B m L )s +
2
(Kb K
+ R Bm )s
(5-123)
and
We
224
Chap. 5
Kb IL a
= e (t)i
~
watts
a (t)
(5-124)
e b {t)ia {t)
hp
746
p
This power
is
K T ddjjt)
~ 146K,m dt
h
K =
t
motor
Tm =
(5-125)
hp
0.737
(5-126)
we have
We
get
hp
P ~ Tm de m (t)
550 dt
curves.
we
(5-127)
0, the
voltage.
speed
Then
is
the back
emf constant
b is
given by
K-a
where in
(5-128)
r
Speed
co m
When
motor
the
T
,
(t).
_
~~
is stalled,
225
rad/sec
designated by
dc motor.
is
Let
(5-129)
K,
= K ijj) = ^E
(5-130)
Also,
{t)
K<
5.8
is
determined:
kR a
(5-131)
Control
phase
motors
more
r^WS
A
motor
6
e l (t)
induction motor.
shown
two
stator with
trical
Reference phase
Fig. 5-43.
in control systems ac
in Fig. 5-43.
The motor
consists of a
degrees apart.
voltage source
reference phase.
is
226
Chap.
is
The control-phase
Speed
cj m
two-phase induc-
tion motor.
The
as follows. Let
blocked-rotor torque at E2
rated control voltage E
E,
_
~~
7\,
k be the
is,
(5-132)
m be a negative number which represents the slope of the linearized torquespeed curve shown in Fig. 5-44. Then
Let
blocked-rotor torque
no-load speed
Tm
(5-133)
fi
is
represented by the
equation
Tm {t) = mmjf) +
where cojf)
is
ke t (t)
(5-134)
voltage.
Now,
if
we
Two-Phase
Sec. 5.8
Induction Motor
may be
227
obtained
from
Jm
^P=-B
n o, m (t)
we have
d9 m {t)
(5-135)
state
is
the
(5-136)
(t)
&>-U-K. + $. M
The
m (t)
= <o m
dt
+ Tm (t)
is
shown
(5-137)
in Fig. 5-45.
The
(m-B m )/Jm
Fig. 5-45. State
placement
is
dis-
obtained as
e m (s)
E 2 (s)
(Bm
m)s[\
+ JJ(B m -
m)s]
(5-138)
or
m (s)
_
s(l
2 (s)
Km
+ r m s)
(5-139)
where
Bm
(5-140)
(5-141)
Since
is a negative number, the equations above show that the effect of
the slope of the torque-speed curve is to add more friction to the motor, thus
stability
228
Chap. 5
of a dc motor. However,
m > Bm
occurs for
5.9
it
if
can be shown
is
is
that the
damping
Step Motors 8
step
motor
is
digital
pulse inputs to analog output shaft motion. In a rotary step motor, the output
rotates in equal increments in response to a train of input
properly
controlled, the output steps of a step motor are equal in
pulses. When
number
to
the
of input pulses. Because modern control systems often
number
shaft of the
motor
have incremental motion of one type or another, step motors have become an
important actuator in recent years. For instance, incremental motion control is
found in all types of computer peripheral equipment, such as printers, tape
variety of
application of a step
motor
mechanism using a
in the paper-drive
step motor.
mechanism of
a printer. In this
revolution,
is
operation.
Sec. 5.9
Step Motors
Stator
229
"C
Rotor "A'
Fig. 5-47.
shown
to have 12 teeth
The motor
rate sections
Phase
Phase
<
'*
<
Phase
<
,,
Fig. 5-49.
End view
230
Sec. 5.9
is
shown
at a position
where
its
stator
To make
tooth pitch on the stator and the rotor are the same.
the
motor
rotate,
the stator sections of the three-phase motor are indexed one-third of a tooth
pitch, in the
same
"Phase
shows
this
Rotor and stator teeth arrangements of a multiple-stack threephase variable-reluctance step motor. The rotor has 10 teeth.
Fig. 5-50.
on one
stator
minimum
of three phases
is
is
and five-phase motors are also common, and motors with as many as eight
phases are available commercially. For an n-phase motor, the stator teeth are
four-
forward. Let any one phase of the windings be energized with a dc signal.
rotor section under the excited phase are aligned opposite the teeth on the excited
phase of the
is
stator.
This
is
in a stable equilibrium.
the position of
minimum
reluctance,
232
If
phase
Chap. 5
is
positioned as shown. It can also be visualized from the same figure that
signal
is
if the dc
switched to phase A, the rotor will rotate by 12, clockwise, and the
A of the stator.
CABCAB
Continu-
motor clock-
Reversing the input sequence will reverse the direction of rotation. That
the input sequence
is,
The
Fig. 5-51.
The nearest rotor tooth axis will always lie within 18 on either side of
The corresponding starting torque exerted when this phase is energized
can be seen in Fig. 5-51. The arrows mark the direction of motion of the rotor.
phase.
this line.
Torque
Fig. 5-51.
step motor.
and
shown
in Fig. 5-50. If
may be
point. This
is
As soon
as phase
A is energized,
inertia
friction are
is
and
is
zero. It
is,
however,
a position of unstable equilibrium since the slightest shift from this position will
0.
point, theoretically
it
to
lie
exactly at the
8
1
in construction,
and the
resulting
asymmetry
We now look upon the stepping motor from a single-step point of view and
try to develop the equations that govern
will
be made
initially to simplify
its
Step Motors
Sec. 5.9
=
R=
L(0) =
i(t)
8(t) =
e(t)
angular displacement
= Ri{t) +
=
start
233
RKt)
stator phase
is
written
[iL(d)]
(5 ' 142)
M
d
+ L{e)f +iji ue)
t
or
e(t)
Ri(t)
L(9)
f+
JJgUp)
(5-143)
The term, L(9)(di/dt) represents transformer electromotive force, or selfinduced electromotive force, and the term i[dL(6)/d9](d6/dt) represents the
back emf due to the rotation of the rotor. We have assumed above that the
inductance is a function of 9(t), the angular displacement, only. No dependence
on the current has been assumed. This will reflect
by the motor.
The energy in the air gap can be written
directly
W = \L{9)i\t)
From
(5-144)
is
we know
given by
T=s [W(i,0)]
where
6{t).
is
(5-145)
i(i)
and
Therefore,
T = \i\t)j
This torque
is
[L{e)]
(5-146)
is
obtained as
T
where Jm
also
is
= jJ^ + Bj-f
Bm
(5-147)
of Eq. (5-146),
we need
to
Jm and Bm
know
the
Z,,
L2
cos rd
(5-148)
234
Chap. 5
r is the
is
(5-149)
Now
let
=
TA =
LA
L,
+L
-Ki\
A is
2
energized.
Then
the inductance
cos rO
sin rd
^K>
eB
=K>
s bB
RB
+ sL B
ebB
~o
Fig. 5-52.
(5-150)
(5-151)
Tension-Control System
Sec. 5.10
235
respectively.
For the
equilibrium position 12
B and C
for phases
=
Lc =
TB =
Tc =
LB
The
electrical circuits
differential
The
- 120)
L,
cos (100 + 120)
-Ki\{t) sin (100 - 120)
-Kil(t) sin (100 + 120)
+L
+L
L,
(5-152)
cos (100
(5-153)
(5-154)
(5-155)
its
total torque
is
the algebraic
sum of
torques
T = TA + TB + Tc
The
(5-156)
motor
made only through computer simulaof the motor, which may be used for
is shown in Fig. 5-52.
tion.
block-diagram representation
analog or
5.10
digital
computer simulation,
Tension-Control System
storage, a roll
wound
it
to explode. Similar examples could be cited, but the need for proper tension
control
is
investigated.
reel.
of the motor.
236
Chap. 5
+ o
Pinch
rolls
is
Armature:
eJLQ
R.Ut)
+ L **p + K co m {t)
a
(5-157)
where
K =
=
b
co m (t)
Torque equation:
TJf)
+ BmeCOm (t) +
^Vnetojt)]
nrT(f)
(5-158)
where
-\-
at
=
BL =
JL
motor
shaft
of windup reel
web material
inertia JL and the
Since the
ceeds, the
effective inertia
is
if
reel
Furthermore,
windup
Jme co m
is
dr
(5-159)
Jr=&>
Thus
dJ,
Ti =
(5-160)
W
i
dr
Tt
where
(5-161)
Sec. 5.11
Assume now
that Hooke's law
that the
is
the
237
C and
a coefficient of elasticity of
obeyed.
is
%p =
where v s {t)
web
C[v w (t)
- ,(/)]
(5-162)
Assuming
It is
is
apparent
now
that because r
thin,
stant.
if
the
web
is
very
~ 0, we may consider that over a certain time period r and JL are con-
%^ =
}
*"
^
5.11
(5-163)
jt-Ut)
*^
me
T aJf) + T
j^cojf)
me
Cnrmjt)
e - (t)
(5 " 164)
"17X0
(5-165)
me
- CV
(5-166)
Whenever there
is
is
roll
may be a dc motor
or a linear actuator.
Figure 5-55 shows the side view of the edge-guide system. The axes of the
1 and 2 are assumed to be fixed or uncontrolled. S^ and S 2 represent
rollers
sensors that are placed at the indicated points to sense the centering of the
at the respective points.
Let
v(t)
^r(0
Z;(0
and
5-55)
web
Roll 2
238
an edge-guide system.
Sec. 5.1
Assuming
that there
is
is
239
roll,
(t):
^2 = t<0tana
(5-167)
at
If the angle
is
small,
Thus Eq.
seen that
it is
~ z (0 - *i(0
(5. 1 68)
m,
dzgt)
v(t)
(5 . 169)
constant, v(t)
is
v,
and Eq.
= *<'>
+ *<'>
Wj
m
at
(5-169) is
(5 " 170)
relation
z.(f)
in 1
of the web
written
z R (t)
ox
^rr\
Z x(j)
Tj = mjv.
Assuming that
= T-r
+ Ti5
1
(5
"
171 )
where
there
is
no
zz(t)
Zl(t)
z R (t)
from
Fig. 5-55,
{t)
(5-172)
or
zz(t)
(t)
+ ^[z
R (t) - z,(0]
m.
n
(5-173)
and solving
for
ZR {s),
Z
Substitution of
tion between
(s)
x
^=
+ \lZ)r,s
and
(5
" 174
>
(5-175)
v
web over
the pivot
roll,
= Zl -T)-
z 3 (t)
where
T=
(t
L {t), sin
on both
Z,(j)
DOAt)
L (t)
is
(5-176)
e- T 'Z t (s)
D6 L (s)
x ,
(5-177)
240
z 3 {t)
and z 5 (t)
Chap. 5
is
Z
Z
where t 3 =
z 3 and z 4 is
3 /v.
5 (s)
3 (s)
(5-178)
r3s
Z 4 (s) =
Z {s)
+ (mjm )z
+rs
the drive
3s
(5-179)
is
an armature-controlled
Us)
where Ia{s)
resistance,
Ejjs)
is
OJs)
(5-180)
= KJJ^s)
Tm(s)
where
- sK
is
(5-181)
TJs)
is
+ Bm s)9 m (s) +
(Jm s 2
LF(s)
(5-182)
where
Jm
Bm =
L
F(s)
inertia of
motor
and
F(s)
= 2^(-^ + B^ s + klWl(s)
2
(5-183)
where
BL
Combining Eqs.
(5-182)
and
(5-183),
BJs)
of
web
we have
1
/me s 2
TJs)
+ B me s + Kme
(5-184)
where
'
me
B,me
"
(5-185)
*~m
\2nr)
K- =
(^)
(4)^
(5-186)
Kl
(5-187)
Also,
X(s)
r9 L {s)
(5-188)
9 L (s)
= *(,)
(5-189)
5-56 using
T3
J*
5
M
241
242
Chap. 5
lers
5.12
is
The blocks
Hp (s) and H
c (s),
the determination
minimized.
Thus
(s)
and
(t)
and z 3 (t)
is
that of a pure
3 (s)
may be encountered
in various types
of systems, especially systems with hydraulic, pneumatic, or mechanical transmissions. In these systems the output will not begin to respond to an input until
Metering
point
S~^
s
(
^>
[*
/I
Valve
Solution
^-\
Solution
(a)
Thickness
measuring gauge
Roller ( o
3)
II
zs
~*
Steel plate
Roller ( o
5)
d
(b)
an arrangement
in
which two
between the mixing and the metering points, the time lag
T= d
-
If it is
is
is
given by
sec
at the
(5-190)
mixing point
is c(t)
and that
it
Sun-Seeker System
Sec. 5.13
measured quantity
243
is
6(0
last
transfer function
between
rollers
(5-191)
is
e~ *C(s)
and
As
(5-192)
c(i) is
~ Ts
in Fig. 5-57(b)
T)
b(t)
m
The arrangement shown
c(t
equation
B{s)
Thus the
(5 - 193)
may be
thought of as a thickness
is
given by Eq.
(5-193).
In terms of state variables, a system with pure time delay can no longer be
described by the matrix-state equation
^p- =
Ax(?)
Bu(0
(5-194)
A general state description of a system containing time lags is given by the following matrix differential-difference equation
^
UL
= t=l
t A,Ht -
77)
+ j=l
Bju(t -
Tj)
(5-195)
where T, and Tj are fixed time delays. In this case Eq. (5-195) represents a
general situation where time delays may exist on both the inputs as well as the
states.
5.13
Sun-Seeker System
In this section
we
shall
is
to
sun with high accuracy. In the system described, tracking of the sun in one plane is accomplished.
schematic diagram of the system is shown in Fig. 5-58. The principal elements
of the error discriminator are two small rectangular silicon photovoltaic cells
it
mination on the
when
the light
is
proportional to the
illu-
244
Chap. 5
Error discriminator
dc tachometer
Fig. 5-58.
voltage,
when
fed to the servoamplifier, will cause the servo to drive the system
is
given as follows.
Coordinate System
The
is
is
and all rotations are measured with respect to this axis. The solar axis or the line
from the output gear to the sun makes an angle d r with respect to the reference
axis, and 9 denotes the vehicle axis with respect to the reference axis. The objeca, near
tive of the control system is to maintain the error between 6 r and
O
,
zero:
6r
(5-196)
When
or Ia
shown
Ib
the vehicle
0.
From
in Fig. 5-58,
is
0.
and Ia
Ib
I,
we have
oa
w + L tan a
= -^^
(5-197)
Ltana
(5-198)
Sun-Seeker System
Sec. 5.13
Center of
output gear
Fig. 5-59.
245
Fixed axis of
dc motor frame
where oa denotes the width of the sun's ray that shines on cell A, and ob is the
same on cell B, for a given a. Since the current Ia is proportional to oa, and Ib
to ob, we have
IT
+ ^tan a
(5-199)
and
I
for
pletely
J
on
cell
A, and
= It
For W/2L
= 21,
Ib
2LI
t
-jp- tan
(5-200)
< tan a < (C - W/2)/L, the sun's ray comFor (C - W/2)/L < tan a < (C + W/2)/L,
= I = for tan a > (C + W/2)/L.
zero.
is
= 0.
may be
teristic
of Fig. 5-60, where for small angle a, tan a has been approximated by a
on the
abscissa.
Fig. 5-60.
is
tan a but
246
Chap. 5
Operational Amplifier
-h-
/.
7f
e
+ -SLjjrI* =
(5-201)
-VWV-
WW-
-JvVW-
1
1Fig. 5-61. Operational amplifier.
Since e
= Aee
eg
= e
/- /
'
If
approaches
Eq. (5-202)
is
becomes
+ (i + i +
infinity, as in
i>
operational amplifiers,
"
(5 202)
A may
reach 10 6
then
written
e
= -R F(I a
(5-203)
I)
Ib
and e
Servoamplifier
The gain of
the servoamplifier
is
is
a.
= -K e
a
With
expressed as
(5-204)
s
Tachometer
the
velocity of
eT
= KT co m
is
related to the
(5-205)
On
9,
(5-206)
00
247
248
Chap. 5
Armature-Controlled DC Motor
The equations
earlier.
are
= RJa + e
e = Kco m
Tm = KJ.
ea
(5-207)
(5-208)
Tm =
where / and
shaft.
B are
J^ +
(5-209)
Bco m
(5-210)
the inertia
The inductance
in the
A block diagram that characterizes all the functional relations of the system
is
shown
in Fig. 5-62.
REFERENCES
State-Variable Analysis of Electric Networks
1.
Kuo, Linear
B. C.
Circuits
New
York, 1967.
2.
Mechanical Systems
3.
New
5.
J.
New
7.
Soc. Amer.
J.,
Vol. 7,
Step Motors
8.
B. C.
Kuo (ed.),
Proceedings,
I,
111.,
1972.
9.
B. C.
Kuo
(ed.),
111.,
1973.
Problems / 249
Chap. 5
10.
B. C.
St.
PROBLEMS
5.1.
Write the force or torque equations for the mechanical systems shown in Fig.
from the force or torque equations.
Af,
B,
M,
F(t)
B,
777777m777m777m7777mm7m7777MM77M7M777,
(a)
*- Fit)
WW///////////////////
(b)
(c)
Figure P5-1.
5.2.
Write a
On
the
set
first try,
2,
0} 2 , 0i,
minimum
order of three.
no,
6l
co
62
Figure P5-2.
cj 2
250
(a)
Chap. 5
Write the state equations in vector-matrix form with the state variables
defined as above.
(b)
Redefine the state variables so that there are only three state equations.
(c)
Draw
(d)
state
results.
(e)
Determine the controllability of the system. Does the fact that the system
can be modeled by three state equations mean that the four-state model is
uncontrollable? Explain.
5.3.
The potentiometer
in Fig. P5-3,
potentiometer terminals
is
(s)/Tm (s).
E volts.
i_
I
Tm it)
Potentiometer
B2 =
Figure P5-3.
5.4.
Write the torque equations of the gear-train system shown in Fig. P5-4. The
moments of inertia of the gears and shafts are Jit J2 and J 3 Tit)
torque. N denotes the number of gear teeth. Assume rigid shafts.
,
is
the applied
Problems
Chap. 5
5.5.
251
Figure P5-5.
(a)
(b)
among
the inputs 0i a
(c)
5.6.
(o s
and
(O x
and (O a
Amplifier
.
eb
Brake
>
Tachometer
(a)
(Sec)
(b)
Figure P5-6.
train
v(t)
Train
is
252
Chap. 5
shown
in Fig.
P5-6b when e b
volt.
Draw
the train.
(c)
of the train
is
to
be maintained at 20
ft/sec,
Figure P5-7.
= applied voltage
R = armature resistance of dc motor
L = armature inductance of dc motor
ea
= armature current
ia
= motor inertia
Bm = motor friction coefficient
JL = inertia of windup reel
co m = angular velocity of dc motor
co = angular velocity of windup reel
TL = torque at the windup reel
r = effective radius of windup reel
V = linear velocity of web at windup reel
T = tension
V = linear velocity of web at input pinch rolls
Jm
Assume
elasticity
of the
web
material
proportional constant
is
is
is
assumed to
satisfy
s,
is
constant.
is,
K (force/displacement).
The
the dis-
and the
Problems
Chap. 5
(a)
i,
co m ,
253
and T as
state variables.
(b)
Assuming
that r
is
constant,
draw a
state
as inputs.
5.8.
Write state equations and output equation for the edge-guide control system
whose block diagram is shown in Fig. 5-56.
5.9.
steel rolling
process
is
shown
in Fig. P5-9.
Two-phase
induction motor
Om
U),Tm {t)
Steel
plate
bit)--
ait)--
Ks dt)
M0
Figure P5-9.
5.10.
TF is
are defined
254
'm
Chap. 5
"m
'm
^_
00
<
^
J
um
e,
G 2 (s)
(s)
>
Integral
control
volt/rad/s
Feedback transducer
(b)
Figure P5-10.
(O m
(a)
Draw
(b)
(c)
is
(e)
When
is
speed
(f)
activated,
TF
is 1
oz-in.
in contact, the
in the steady-state
10 V.
torque 7>
it
shown by
is
suggested that a
Problems
Chap. 5
and the
5.11.
motor speed
is
compared with
Draw
(g)
The
255
First
When traveling
control standpoint
missile
about
its
is
air,
is
C is traveling, as shown in
Fig. P5-1
Figure P5-11.
with the angle 6 (9 is also called the angle of attack), a side force is produced by
the resistance of the air through which the missile is traveling. The total force
Fa may
Normally, a F
is
a, F
where a
is
force, be denoted by
of attack 6 and is given by
Ct F .
= ad
a constant described by
Kr
=
d =
J
t
The main
missile
moment
of inertia about
distance between
C and P
is
control
means
is
to
256
Chap. 5
tail
the system
parameters.
(b)
Assume
that
T is
S.
5.12.
(c)
Repeat
(a)
Draw
(a)
and
(b)
C and P interchanged.
a state diagram for the tension-control system of Fig. 5-53, using the
(b)
and
(5-166).
EJs) and
T(s), with
as inputs
and
system
x(t)
i,(t)
=
=
v{t)
-k(v)
- g(x) +
T(t)
where
x(t)
v(t)
k{v)
g{x)
T{t)
The
electric
motor
is
relations
= Kb ^>(t)v{t) +
= Km <f>{t)ia (t)
e(t)
T(t)
Ria {t)
where
R =
armature resistance
<j>(t)
(a)
= Bv(t), and R =
Consider that
system. g{x)
(c)
is
x(t)
(b)
0.
ia (t)
0,
Consider that
f (t)
k(v)
<f>(t) is
v(t)
is
the input
= Bv(t).
0, k(v) =
Bv{t),
state
maximized.
Repeat part
(a)
when
TL
is
zero.
<X L ,
is
Chap. 5
Problems
257
N,
=
\
Jl
V
1
/
4
V
N, EE
Figure P5-14.
5.15.
(a)
j-'k8
in Fig.
7
%
Figure P5-15.
(b)
Show
of the form
= Ax
where
o
--L!.i
A=r_-j-'k o_
;
(c)
5.16.
set
K2 =
3000, Ji
1,
J2
5,
J3
2,
and
K =
3
A",
1000,
1000. Find the matrix A.
Figure P5-16 shows the layout of the control of the unwind process of a cable
reel with the object of maintaining constant linear cable velocity. Control
is
established by measuring the cable velocity,
signal,
comparing
signal.
it
with a reference
A tachometer is used to
sense the cable velocity. To maintain a constant linear cable velocity, the
angular reel velocity 9 R must increase as the cable unwinds; that is, as the
effective radius of the reel decreases. Let
D=
cable diameter
W = width of reel
jR
= 0.1
2
ft
(empty
reel)
2 f
258
Chap. 5
v:
'OrIM^
Motor
Tachometer
+
em
Amplifier
Figure P5-16.
Rf =
R=
effective radius
of reel
effective radius
of reel
(full reel)
200
ft-lb-sec 2
vR
Motor
inertia
and
The tachometer
E,{s)
VR(s)
and the motor
transfer function
transfer function
is
+ 0.5s
is
Tm (s)
Em {s)
50
s
(c)
VR (s)/E (s).
r
6
Time-Domain Analysis
of Control
6.1
Systems
Introduction
Since time
is used as an independent variable in most control systems, it is usuof interest to evaluate the time response of the systems. In the analysis
problem, a reference input signal is applied to a system, and the performance of
the system is evaluated by studying the response in the time domain. For
ally
is
it is
it
may
c(t)
divided, into
two
parts:
be written
c,(t)
c(t)
(6-1)
where
= transient response
c n(t) = steady-state response
c,(t)
The
analysis
it is
is
is considered as a steady-state response because its behavany time interval, as when time approaches infinity. Similarly, if
described by c(t) = t, it may be defined as a steady-state response.
fixed for
a response
is
259
260
Time-Domain Analysis
Chap. 6
of Control Systems
Transient response
as time
becomes
is
lim
c,(f)
(6-2)
is
phenomenon
to
some
extent before a
steady state
The
is
is
of importance, since
it is
part
of the dynamic behavior of the system; and the deviation between the response
and the input or the desired response, before the steady state is reached, must
be closely watched. The steady-state response, when compared with the input,
gives an indication of the final accuracy of the system. If the steady-state
response of the output does not agree with the steady state of the input exactly,
the system
6.2
is
to time.
For
vary in
lem for the designer, since it is difficult to design the control system so that it
will perform satisfactorily to any input signal. For the purposes of analysis and
design, it is necessary to assume some basic types of input functions so that the
performance of a system can be evaluated with respect to these test signals. By
selecting these basic test signals properly, not only the mathematical treatment
of the problem is systematized, but the responses due to these inputs allow the
prediction of the system's performance to other more complex inputs. In a
design problem, performance criteria may be specified with respect to these test
When
Sec. 6.2
261
position of a mechanical shaft, the step input represents the sudden rotation of
the shaft.
where
R is
a constant.
The
is
(6-3)
Or
r{t)
where u s (t)
is
>0
<0
K0 =
step function
Ru
The
is
(6-4)
s (t)
step function
shown
is
not defined at
0.
in Fig. 6-l(a).
(b)
(a)
time-domain
Ramp
= Ru
test signals
s (t).
(b)
Ramp
Rtu s {t).
Rt 2 u s (t).
ramp
is
con-
ramp
function
is
represented by
K0 =
t>0
."
?<0
\Rt
L
[0
(6
-5
or simply
/(f)
is
shown
Rtu s (0
(6-6)
ramp input
form of
input function
is
, ,
\Rt 2
>
262
Time-Domain Analysis
of Control Systems
Chap. 6
or simply
<0 =
The
Rt*ut)
(6-8)
These
scribe mathematically,
become progressively
as a test signal since
have the
common
is
shown
in Fig. 6-l(c).
instantaneous
its initial
deal about a system's quickness to respond. Also, since the step function has, in
principle, a
it
its
is
how
the system
would respond to
than a ramp function. In practice, we seldom find it necessary to use a test signal
faster than a parabolic function. This is because, as we shall show later, to track
or follow a high-order input, the system is necessarily of high order, which may
mean that stability problems will be encountered.
6.3
we
system
may be
response or,
as an
following.
Steady-State Error
It
was mentioned
accuracy
when a
specific type
of input
is
measure of system
sical system, because of friction and the nature of the particular system, the
steady state of the output response seldom agrees exactly with the reference
input. Therefore, steady-state errors in control systems are almost unavoidable,
and in a design problem one of the objectives is to keep the error to a minimum
or below a certain tolerable value. For instance, in a positional control system,
it is desirable to have the final position of the output be in exact correspondence
with the reference position. In a velocity-control system, the objective is to have
the output velocity be as close as possible to the reference value.
If the reference input r{t) and the controlled output c(t) are dimensionally
the same, for example, a voltage controlling a voltage, a position controlling a
position,
is
are at the
same
level
simply
e(t)
= r(t) -
c{t)
(6-9)
Time-Domain Performance
Sec. 6.3
However, sometimes
reference input that
may be
it
at the
is
same
of Control
Systems
263
level or
may
it
source for the control of the output of a high-voltage power source; for a
velocity-control system
it is
more
Under
error signal cannot be defined simply as the difference between the reference
input and the controlled output, and Eq. (6-9) becomes meaningless. The input
signals
/\
r(t)
6-2.
The
is
at the
same
level
usually incorporated
e(0
cit)
G(s)
>
as)
R(s)
bit)
H(s)
Bis)
Fig. 6-2.
control system
is
defined as
e (0
r{t)
&{s)
R(s)
b(t)
(6-10)
or
For example,
is
if
a 10-v reference
a constant and
error signal
is
equal to
0.1.
is
B(s)
R(s)
H(s)C(s)
(6-1 1)
When
is
exactly 100
the
is
f(0= 10-0.1-100 =
As another example
let
(6-12)
shown
is a
used as a reference to control the
we need a
v,
c(t)
is
in Fig. 6-2
in velocity
is
K,s.
defined as
e (0
r{t)
b{t)
r{t)
k;-
(6-13)
dc(t)
dt
The error becomes zero when the output velocity dc(i)\dt is equal to r(i)\K
The steady-state error of a feedback control system is defined as the error
when time reaches infinity; that is,
t
steady-state error
lim
e(t)
(6-14)
264
Time-Domain Analysis
of Control Systems
With reference
Chap. 6
By use of the
s&(s) is to
^ = r+W)W)
(6 ' 15)
where
is
lim
e(t)
lie
lim s&(s)
(6-16)
on the imaginary
lim
*fffr).
,
axis
and
we have
is
(6-17)
>
which shows that the steady-state error depends on the reference input R(s)
and the loop transfer function G(s)H(s).
Let us first establish the type of control system by referring to the form of
G(s)H(s). In general, G(s)H(s)
G(s)H(s)
{)
may
be written
+ T s)(l + T s)
s'(l
(1
s)
(6m
{t 1V)
+ T s)
n
K and all the Ts are constants. The type of feedback control system refers
= 0. Therefore, the system that is described by the G(s)H(s) o(Eq. (6-1 8) is of type/, wherey = 0, 1,2, ... The values
where
and the Ts are not important to the system type and do not affect the
value of the steady-state error. For instance, a feedback control system with
of m,
n,
G <*M*>
is
of type
1,
since j
= /+^(1+1)
(6 " 19)
1.
steady-state
Steady-state error due to a step input. If the reference input to the control
is a step input of magnitude R, the Laplace transform of the
Equation (6-17) becomes
is
R/s.
sR(s)
i:
"
+ G(s)H(s)
~~
G(s)H(s)
~~
1
R_
lim G(s)H(s)
fi
?m
'
Kp =
where K
is
We
infinite. If
lim G(s)H(s)
G(s)H(s)
is
"
(6-21)
Then Eq.
(6-20)
is
written
= TTK,
when
(6 " 22)
the input
is
we
a step function,
see that for
Kp
must be
to be infinite,
j must be at least equal to unity; that is, G(s)H(s) must have at least one pure
integration. Therefore, we can summarize the steady-state error due to a step
Time-Domain Performance
Sec. 6.3
of Control
Systems
265
input as follows:
system
type
type
-"
of Fig. 6-2
ramp
is
K,
constant
system
is
r(t)
where
e ss
~ +
=
(6-23)
Rtu,(t)
r(t) is
(6-24)
If
we
Mm
ss
we have
V"o s
sG(s)H(s)
(6-25)
lim sG(s)H(s)
define
K =
v
lim sG(s)H(s)
(6-26)
K.
is
ramp
function.
typical e s
e ss
Reference input
r(l)
R/Kv
= Rtu s (t)
is
ramp
function,
K =
v
lim sG(s)H(s)
s-0
lim
J
s-0 S
to be infinite, j
j=
must be
0,1,2,.
(6-28)
may be
stated
266
Time-Domain Analysis
of Control Systems
Chap. 6
system
e ss
oo
system:
e ss
--
e ss
constant
= \u
2
described by
(6-29)
s {t)
r(t) is
= -J
R(s)
The
is
(6-30)
is
R
"
(6-31)
lim s 2 G{s)H(s)
K =
a
lim s 2 G(s)H(s)
(6-32)
is
e ss
(6-33)
= constant
e ss
system:
type 2 system:
oo
e ss
type
As
e ss
=
=
system:
type
oo
among the
the relations
error constants,
the types of the system, and the input types are tabulated in Table 6-1.
transfer function of Eq. (6-18)
Table 6-1
Summary
is
used as a reference.
Due
to Step,
Ramp,
Type of
System
-is
Ap
js
Ay
v-
Aa
&ss
Step
Ramp
Input,
Input,
if
1 ~r Jy p
\
oo
oo
oo
e ss
oo
oo
oo
e ss
=
=
=
_ R
is
Input,
^ SS
_ R
T^~
A
a
A;,
"
=
1
^**
Parabolic
+K
e ss
e ss
e 3S
=
=
=
^
e ss
The
Time-Domain Performance
Sec. 6.3
of Control
Systems
267
should be noted that the position, velocity, and acceleration error constants are significant in the error analysis only when the input signal is a step
function, a ramp function, and a parabolic function, respectively.
It
It
should be noted further that the steady-state error analysis in this section
is
e(t)
lit)
(6-34)
c(t)
Then
()
G(s)[H(s)
l]
R()
(6 . 35)
and
e
It
here
-lims l
G(s)[H(s)
l]
(6-36)
R(s)
should be kept in mind that since the steady-state error analysis discussed
on the use of the final-value theorem, it is important to first check to
relies
One
is
is
that
when
the steady-state
We shall
Error Series
In this section, the error-constant concept
We
is
start
S ^>
r+Mm
(6 " 37)
may be
(r)
where
we (r)
is
3.3,
written
J" _
w e (t)r{t
- t) dx
(6-38)
W^= +G \s)H(s)
<
W9
>
which
is
known
If the first
all
values of
/,
t)
268
Time-Domain Analysis
of Control Systems
Chap. 6
where
T)
series; that
r(t)
xr(t)
is,
- ^r(t) +
|if(/)
...
(6-40)
represents the
r(t)
Since
r(t) is
we have
w.(T)[r(0
+ ff (0 - yj'C) +
*H?)
.]
rfr
(6-41)
w e (r) dx
lit)
As
xw e (x) dx
r(t)
f ^w.(i) dx
r(t)
approaches
is
= lim e(t) =
lim
w e (x)dx-f
r,(t)\
xw e (x)dx
(t)\
(6-42)
e,(f)
e(t) as
thus
infinity;
is
given by
r s (t)\
~^\{x)dx
(6-43)
and
r,(t)
^(0
^w,(x)dz
r(t).
Let us define
C =
{x)
dx
Ci
xwx) dx
J
C2 =
Equation (6-42)
e,(0
which
is
=C
is
x 2 w e (x) dx
(6-44)
= (-l)TT"W
(TVT
J
e
written
r,(t)
C,r,(0
+ ^(f) +
and the
+ ^r(/) +
coefficients,
C C C2
,
];
(6-45)
C are
The
error coefficients
may be
transform,
we have
W. is)
= P w,(t>-" rft
J
(6-46)
Time-Domain Performance
Sec. 6.3
Taking the
limit
on both
of Control
lim
e (s)
269
we have
f w (T)e"" dx
lim
i-o
s^o
Systems
>
(6-47)
=C
The
derivative of
Ws) of Eq.
-f^ =
ds
Tw e (r)e " di
Jo
(6-48)
= de'"
from which we
get
C,
The
rest
lim
4EM-
.-o
ds
(6-49)
Cz =
3
(6-50)
ds 2
lim)
ds
(6-51)
v
*^o
and
Therefore,
^M
lim
,^o
C =
s.
of the error
series
Example
6-1
ficients.
G(s)
Thus the
is
of type
= j^-j
(6-53)
= K, Kv = 0, and a = 0.
p
due to the three basic types of inputs are as
follows
unit step input, u,(t):
e ss
unit
ramp
input, tu s (t)
Notice that
state error
is
when
infinite in
the input
is
either a
magnitude, since
it
e ss
2
u s (t):
e ss
ramp or a
=
=
oo
oo
manner
in
It is
apparent
actually be determined
series.
270
Time-Domain Analysis
of Control Systems
Using Eq.
(6-39),
Chap. 6
we have for
system
this
(6 " 54)
C =
= Kjr-x-f
+
lim W.(s)
(6
j^.0
<=il-T = (rTTF
C = ~d^- = + *)3
Although higher-order
Now
let
(6 " 57)
(1
coefficients
The
55 >
(6 " 56)
>
= YTK r*W +
W+
+K)*
(1
become
will
error series
"''
+^)
(1
less significant
written
is
>
(?)
(6 " 58)
When
tives
of
r,(f)
is
The error
are zero.
*,(0
When
and
result given
(t),
and
all
deriva-
(6-59)
all
=u
r s (t)
series gives
is
is
- (0
TTT-F
+ K t'
_1
+ TTZT^I
(1 +KY
all
fails
linearly
60)
with time.
input, rt)
(t
tu s (t), r s (t)
u s {t), and
4.
but
infinite
is
For a parabolic
(6
-(')
(6 " 61)
where a
[o
+a +
t
^+
e-~jujf)
(6-62)
r,(f)
fit)
rXt)
+ <M + ^],(0
[a
= (i + a 2 t)u
= a 2 u (t)
(6-63)
(6-64)
s (t)
(6-65)
=
1
+ K n(t) +
(1
+K)
M-
K
(i
+^)3 ^(f)
(6
" 66
>
Time-Domain Performance
Sec. 6.4
Example
In this example
6-2
stant
is
we
shall consider
6-1
is
(D
sin (D
(6-67)
= 2. Then
= sin (D
= (D cos (D
r,(t) = -Oil sin (D
r (t) = -G>1 cos (D
r,(i)
rjit)
The
e,(t)
a sinusoid,
r(t)
where
(6-68)
t
= [c -
2y C0
4T o
sin
V +
[Cio
-yfcoS
cos (D
(6-69)
Because of the sinusoidal input, the error series is now an infinite series. The conis important in arriving at a meaningful answer to the steady-
series
C =
Cl
YTH = 00099
= (1 +K? = 00098
c* = ~ (i
K
+Ky = 0000194
6K
"
C = (i
+
, =
xy
3
first
x 10" 8
et) =i ro.0099
L
5.65
0.01029 sin It
(6
70 )
0.0196 cos It
or
e,(0
6.4
is
(6-71)
by Eq.
(6-71).
The
is
is
Of course,
is
when
referred to, since for an unstable system the response does not
out of control.
272
Time-Domain Analysis
of Control Systems
The
Chap. 6
is
usually characterized by
the use of a unit step input. Typical performance criteria thai are used to characterize the transient response to
rise time,
and
settling time.
Maximum
overshoot
with respect to
Maximum
overshoot.
is
defined as the
largest deviation of the output over the step input during the transient state.
measure of the
shoot
is
per cent
maximum
is,
overshoot
= maximum overshoot
final
value
X 100%
(6-72)
2.
Td is
Rise time.
The
rise
time
is
its final
value.
its final
value.
Sec. 6.5
273
reciprocal of the slope of the step response at the instant that the
response
is
Settling time.
4.
The
settling
time
its final
is
value.
of its
The four
final value.
when
sure
tities
5 per cent.
a step response
is
relatively easy to
mea-
Performance Index
Since the general design objective of a control system is to have a small
overshoot, fast rise time, short delay time, short settling time, and low steadyit is advantageous to use a performance index that gives a measure
of the overall quality of the response. Let us define the input signal of a system
as r(t) and the output as c(t). The difference between the input and the output
state error,
is
r(t) is
referred to as
In trying to minimize the error signal, time integrals of functions of the error
may be used as performance indices. For example, the simplest integral
signal
is
dt
(6-73)
where /
is
thus a zero
/.
Some of the
\"\e{f)\dt
J o
6.5
others.
performance indices
is
[te{t)dt
Jo
f e\t)dt
JO
Although true second-order control systems are rare in practice, their analysis
generally helps to form a basis for the understanding of design and analysis
techniques.
state
Ji(t)J
where and
co
\_-<ol
are constants.
-2Cco,
"*i(0"
r(t)
(6-74)
274
Time-Domain Analysis
of Control
Systems
Chap. 6
x 2 (0+)/s
x,(0+)/s
is
c(0
colx^t)
state
(6-75)
diagram of
's
2Cco
-col
*i(0+)"
R(s)
(6-76)
s_\ix 2 (0+)
where
A=
The
2cos
For a
sin (co*s/l
*l(0
Vi -C
2
1
co sin g>V
1
col
yi-c
w*J\
=e"
fo "'
is
we have
y/)
sin
co n
(6-77)
col
x 2 (t) J
s2
sin
awT-
(eoV 1
e^-'sinKVl -C 2 '-
*i(0+)
$)
.* 2 (0+)J
(6-78)
<t>\\
r>0
sin
co,,
VI -C 2
'
where
y/
tan
yi-c
(6-79)
=
Although Eq.
terms of the
tan
-,yi-c
(6-80)
initial states
it is
a rather formidable-looking
is
order. However, the analysis of control systems does not rely completely on the
The development of
performance by use of
275
Sec. 6.5
and the
characteristic equation.
a great
deal can be learned about the system's behavior by studying the location of the
roots of the characteristic equation.
col
R(s)
The
zero
that
(6
K
+ col
2gv
Fig. 6-5.
. 81
)
}
is
col
is,
A=
2Ccos
(6-82)
is
C(s)
=
s{s
Or,
c(t) is
f'
2cos
(6-83)
+ col)
= +
c(0
-CoJnf
It is interesting to
co*/l
7r=c sSin
tan -,-v/l-C
-c
initial states
2
t
>
(6-84)
c(t).
The two
are
su
The
s2
=
=
-C
(0-83)
a.jco
and
, co n , a,
co is
now
described
as follows:
As
multiplied to
in the exponential
rate of rise
factor.
The
When
inverse of a, 1/a,
is
the two roots of the characteristic equation are real and identical
From
we
we
damping
occurs when f = 1. Under this condition the damping factor is simply a = co.
Therefore, we can regard as the damping ratio, which is the ratio between the
actual damping factor and the damping factor when the damping is critical.
co is defined as the natural undamped frequency. As seen from Eq. (6-85),
call the
system
critically
damped.
Eq. (6-85)
when
0,
= aw^ -
(6-86)
the response of Eq. (6-84) is not a periodic funcnot a frequency. For the purpose of reference co is
276
Time-Domain Analysis
of Control
Systems
Chap. 6
l'u
Root
s-plane
)
;,
= >/!--r 2
\
N
"
*-a =
$u
Root
Figure 6-6 illustrates the relationship between the location of the characteristic
shown,
a> is
C> cu,
and
ca.
the radial distance from the roots to the origin of the s-plane.
damping factor a is the real part of the roots; the conditional frequency is the
imaginary part of the roots, and the damping ratio is equal to the cosine of
the angle between the radial line to the roots and the negative real axis; that
cos 9
loci,
is,
(6-87)
the constant-^
loci,
the constant-a
and the constant-cu loci. Note that the left-half of the j-plane corresponds to positive damping (i.e., the damping factor or ratio is positive), and the
right-half of the j-plane corresponds to negative damping. The imaginary axis
corresponds to zero damping (a = 0, = 0). As shown by Eq. (6-84), when
the damping is positive, the step response will settle to its constant final value
because of the negative exponent of e~ Ca*. Negative damping will correspond
to a response that grows without bound, and zero damping gives rise to a sustained sinusoidal oscillation. These last two cases are defined as unstable for
loci,
equation roots plays a great part in the dynamic behavior of the transient
Sec. 6.5
s-plane
\
^ \\
XyA J
'
/to
277
s-plane
f = o
Negative
>N
Positive
damping
damping
Positive
damping
'
/]
1
Negative
damping
f2>fl
r = o
(b)
/CO
/OJ
,
s-plane
s-plane
Positive
Negative
damping
damping
co 2
00,
-2
-t-
-Oij
-3
-to,
Positive
Negative
damping
damping
a2
-co 2
>a >0
l
<*
(d)
(c)
damping
order system
<C<
s u s2
-Ccon
j(O *J\~
n
underdamped case
critically
C>i
C = o
Si, s 2
C<o
damped
case
overdamped case
t ,
s2
= /cu
= Ca> jco^/l
undamped
case
negatively
damped
case
/oo
x-plane
r=o
_^
<?<!/'
0>f>-l
?=
r>i
*-
5l
f<-I
t-
f-
yo>r>-i
r
Fig. 6-8.
damping
=o
a> is
/"
c(,t)
s-plane
*-
-X
f>l
/OJ
x-plane
f=l
x-plane
/
c(/)
o<r<i
*-f
X
Fig. 6-9.
278
in the x-plane.
Sec. 6.5
279
c(t)
/co
s-plane
>~ct
f =
/CO
s-plane
0>f>-l
/CO
s-plane
r<-i
locations.
>
Thus
dc(t)
tcoe
-Zcon t
sin (cot
dt
t4)
(6-88)
-CcOnt
~
/i
_^
CQa
^/1
cos (a>t
4>)
t>0
280
Time-Domain Analysis
of Control
Chap. 6
Systems
Fig. 6-10. Transient response of a second-order system to a unit step function input.
where
(j>
Equation (6-88)
dc(t)
dt
is
tan
-,
yi-c
(6-89)
simplified to
co
Jl -
e
"
-&,<
cVl-C 2 ' =
si n
q^ ^i
we have
2,
oo
>
(6-90)
and
0,1,2,...
(6-91)
Sec. 6.5
281
or
nn
1.
Therefore,
(6-93)
..,
/,
c(t)
v7~p vT^p"
vT^T1
vT^F
at
<\0
Im.x or
mm
==- sin (
r
Vi
-C 2
,
nn
tan"
^ZI^-)
[
il
"=1,2,3,...
(6-94)
or
C(OIorml.
= +
1
(-I)""'* "f=F
maximum
overshoot
maximum
overshoot
obtained by letting n
and
per cent
(6-95)
=
= c max
is
lOOer*^ 1
e -'C/-/T=c
(6-96)
(6-97)
282
Time-Domain Analysis
of Control
Chap. 6
Systems
0.8
0.6
0.4
Damping
1.0
1.2
ratio f
damping
Note that
is
per cent
maximum
is
shown
maximum
response
in Fig. 6-12.
From
consideration, the
maximum
it is
it
occurs are
all
and
co.
^~l+0.7
Thus the delay time
(6-98)
is
+ 0-7C
1
(6-99)
co
id
_1 + 0.6 + 0.15
=
(6-100)
CO
For the
rise
time
tr ,
which
is
10 per cent to 90 per cent of its final value, the exact values can again be obtained
directly
Fig. 6-10.
The
plot of
by a
cot r
versus
is
shown
in
0.8
2.5
o<<
(6-101)
COn
better approximation
may be
Sec. 6.5
283
= l+0.7f
H'rf
<
<%) _
R (s)
0.5
0.2
0.4
0.6
0.8
s2+2$us +
1.0
1.2
w n2
1.4
1.6
f
Fig. 6-13.
system.
5.0
"i
4.0
2.0
-I
0.2
0.4
0.6
J_
0.8
1.0
1.2
?
Fig. 6-14.
Normalized
rise
then
~ +
l
,_
l-K+lAC 2
co
From
shown
it is
(6-102)
<f<
in Fig. 6-15.
284
Fig. 6-15.
Chap. 6
From
the figure
it is
same
<
<
result is
1).
Solving for
cot s
from the
is
used. Therefore,
-CcOnl.
1.05
last equation,
(6-103)
we have
<oj,= -4-ln[0.05yi-C]
For small values of f, Eq. (6-104)
CO n t s
is
(6-104)
simplified to
= -y
(6-105)
or
s
Now
- c<
o<c<i
(6-106)
rise time,
and
settling
seen that small values of f would yield short rise time and short delay
time. However, a fast settling time requires a large value for . Therefore, a
compromise in the value of should be made when all these criteria are to be
time,
it is
satisfactorily
met
consideration on
maximum
overshoot, a generally accepted range of damping ratio for satisfactory all-around performance is between 0.5 and 0.8.
6.6
Time Response of
a Positional Control
System
Sec. 6.6
285
dc motor
dc
amplifier
if
= constant
*
Error detector
t ^Q09
set
of potentiometers
dcmot
TH current
,!
a dc
motor. The
the
field of the dc motor is held
constant
The parameters of the system are given
as follows:
K = 1/57.3 volt/deg =
A (variable)
* = 5Q
L = negligible
Jm = 10" lb-ft-sec*
Bm = negligible
BL = 0.1 lb-ft-sec
= 0.11b-ft-sec
= NJN 1 =
TV
K = 0.5 lb-ft/amp
s
Gain of dc amplifier
Resistance of armature of motor
Inductance of armature of motor
of rotor of motor
Friction of motor shaft
Inertia
volt/rad
/z,
Gear ratio
Torque constant of motor
The
first
is
1.
Error detector:
9eQ)
e(t)
2.
This
cause-
= e (t) - exo
= KftJit)
r
DC amplifier:
ea {t)
= Ae(t)
(6-107)
(6-108)
(6-109)
286
Time-Domain Analysis
of Control
3.
Chap. 6
Systems
Armature-controlled dc motor:
LjMl = - RJ {t) +
e a (t)
e{t)
where
b is
= K co m {t)
(6-111)
the back
TJf)
Km
(6
jj-^- = -Bme co m +
(t)
Bme
= Jm +
Bme = Bm
Jme
4.
112 )
(6-113)
TJt)
respectively,
= lO' + 0.01(0.1) = 2
n*BL = 10" lb-ft-sec
n 2 JL
-\-
by the motor,
coefficients seen
(6-1 10)
e b (t)
and
10- 3 lb-ft-sec 2
Output:
(6-114)
(6-1
5)
co m {t)
(6-116)
n8Jt)
(6-117)
0c(t)
The value of the back emf constant, Kb is not given originally, but a definite
relationship exists between Kb and K In the British unit system, K, is given in
lb-ft/amp and the units of the back emf constant are volts/rad/sec. With these
units, Kb is related to K, through a constant ratio. The mechanical power
,
P(f)
Also,
it is
known
is
e b {t)i a {t)
^e
and
watts
(6-118)
b (t)it)
hp
= ^TJfyoJt)
hp
we have
(6-119)
that
Pit)
= -^Tjfajt)
hp
(6-120)
A,
= 2a =
/40
0.737A.
(6-121)
or
K =
b
1.36*;
(6-122)
form
as follows:
Sec. 6.6
Time Response
dia {ty
of a Positional Control
nAK,
dco m (t)
Bme
dt
J me
co m (t)
'me
287
[AK.1
r us)
dt
System
0,(0
(6-123)
dBJf)
dt
is
is
[0(O
Fig. 6-17.
9<a (0+)A
9 B m (0
+)/s
"AK.IL.
Fig. 6-17. State
When
is
assumed to be
obtained as
K AK n
0,(0
in Fig. 6-16.
*A.4l + V>U +
W) + K K
b
lS
+ K.AK,n
(6 " 124)
where
?.
^ me
The
~n
0.(0
The open-loop
negligible
^ SeC
= JsW =
is
denned
as
0,(0-0,(0
G(s)
The
is
(6-125)
K,AK n
K ^meS{l +
t.j)(1
r me s)
+KK
b
(6-126)
lS
state
288
Time-Domain Analysis
Chap. 6
Systems
of Control
the positional control system with respect to the system parameters, and
sufficient to assume that all the initial states of the system are zero.
Since
(6-124)
is
La
is
negligible, t
0,
it
is
simplified to
K,AK,n
OJs)
R.Jm
d,{s)
y + (K K
t
+ R aB m .)s +
(&l27)
v
K,AK n
l
Equation (6-127)
is
AK,n
w =
i V KR.Jm
s
(6-128)
v
"
The damping
ratio
is
_ K K, + RB me _ K K, + R B m
2jK AK,RJme n
2R Jme(o
a
When
(6-129)
the values of the system parameters are substituted into Eq. (6-127), the
JA -
5A
VJs)-s 1 +
34.5s
damping
the
=
C =
The
The natural
31.6rad/sec
(6-131)
0.546
(6-132)
34.5s
(6-130)
'
l
5A
is
1000
(6-133)
-17.25 j26.5
(6-134)
^=
'
or
fl
rA =-
4-
1000
+ 3^7+
9 <,-o.5 4 6<r
1.2e-
17
"'
(6 " ,35)
1000) J
^ n rn.R17m
sin (26.4?
is
u s (t) rad;
t -i-
ir.
-i-
tan"
1.53)
(6-136)
236.8)
cot.
It is interesting to
when
the value of
the gain
the gain
Sec. 6.6
Time Response
of a Positional Control
System
289
Bc {t)
Fig. 6-18.
the input
in Fig. 6-16
when
When A
is set
at a relatively
undamped frequency
m =
Table 6-2
Damping
Gain
13.5
200
1500
Ratio C
2.1
Maximum
(On
Overshoot
8.22
Tt
Tr
Ts
'max
0.325
1.02
1.51
0.546
31.6
0.13
0.043
0.064
0.174
0.2
0.119
86.6
0.52
0.013
0.015
0.17
0.037
290
Time-Domain Analysis
of Control
When
Chap. 6
Systems
the value of
is set
and
A =
at 13.5
13.5
=
=
s u s2
,4=200
as
characteristic
listed as follows:
jj 2
-2.1, -32.4
-17.25 y26.
These roots, together with the ones for A = 1500, are located in the s-plane
in Fig. 6-19. In general, for any value of A, the roots of the character-
shown
teristic
,s 2
-17.25
1
-j-*/U90-20A
/86
= 1500
(6-137)
.s-plane
^
<-
A<0
,4
-2.1
= 135
.4
= 200
;26.5
A =
= 13.5
A<0
-/26.5
Root
A = 1500
A
Fig. 6-19.
A between
loci
/86
lie
and the
When A
is
is
is
all
positive,
eC
'
Time Response
'
of a Positional Control
System
291
Amplifier Gain
<A<
A
59.5
System Dynamics
Two
Overdamped
Two
59.5
59.5
-oo
<A <
Two complex
oo
conjugate roots
with negative real parts
<A<
Two
positive
one
and one negative
>
1)
damped
Critically
roots
= n
Underdamped
<
1)
Unstable system
(f
< 0)
K
v
j>
"i? 5(s
is
The
(6
34.5)
'
138 )
= T+X =
( 6 -139)
Time Response to a
When
tem
5A
= lim
a unit
Ramp input
ramp function
input 0,(0
0,(0
From
JB-'
is
described by
UV +
2Cto.fi
U0 = ~ S +
aWl'-C^
'
'
is
(6-140)
+ coi)
written
Sin [(a
^TZ=T
2
'
(6-H1)
where
= 2tan->Vl^
= 115 20' and 150 are sketched as
he
nSeS f r
P re
'
f,
Fig.
6-20
Notice that in this case the steady-state error
of the system
As seen from Eq. (6-141), the last term will decay
to zero as
infinity.
nT
.
(6 142)
shown
in
nonzero
time approaches
Therefore, the steady-state response
of the system due to a unit ramp
is
292
Time-Domain Analysis
Chap. 6
of Control Systems
4.0
e(A = 200)
ea (A
3.0
= 1500) s
= 1500
e ss
0,(0 = tu s (t)
2.0
Fig. 6-20.
when
input
= 13.5
= 200
the input
is
a unit
ramp
(A = 13.5)
in Fig. 6-16
function.
is
lim 6 c {t)
_2f
lim
(6-143)
<w.
It is
34.5
is
is
(6-144)
5A
a>
which
ramp input
a constant.
error to a
ramp
v.
K,
lim sG(s)H(s)
lim
^3 = ^
(6-145)
result
34.5
(6-146)
of Eq. (6-144).
more
oscillatory.
This phenomenon
higher-order systems,
become
unstable.
if
is
rather typical in
is
all
may
Time Response
of a Positional Control
System
293
was shown
dc motor
is
constant To
(6-124)
is
now
is
The closed-loop
0.02 sec.
now
&)
9 r (s)
__
armature time
by
y Eq
4
0.05,4
U.005<1
+ 2s) +
0.02j)(1
34s
or
js
1725^
s(s>
50.5,
50.5s 2
\725s
ris)
The
(6
1725)
"
148 )
"
149 )
is
250A
is
(6-150)
G)
at
(6
250,4
flcfr)
~W)~
is
0.05,4
250,4
It is
50.5**
transfer function
G (A
{S)
The
250,4
*(*)
0M
The open-loop
'
now
(1
0.48 j)(l
0.0298j
0.000616* 2 )
(6_151 )
fl e
(0
= -
I.06e-""
0.0667e--* sin(32.2<
1.88)
(6-152)
In Eq. (6-1 52), since the time constant
of the pure exponential term is more
than 10 times greater than that of the damped
sinusoid term, the transient response of 8t) is predominantly governed
by the
1
characteristic root at s
-2.08. The response due to the last term of
Eq. (6-152) decreases to zero very
rapidly with the increase of t. Comparing
Eq. (6-152) with Eq. (6-136) we see
that the damping factor for the second-order
system (L
0) is 17 25 whereas
for the third-order system (L.
0.1) the output response is governed by
the
exponential term, which has an equivalent
damping factor of 2.08 Thus the
third-order system will have a slower rise
time. This result is expected, since
the presence of inductance will slow
down the buildup of the armature current
thus slowing down the torque development
of the motor. However, higher inductance will cause a higher overshoot in the
step response, as shown in Fig 6-21
With
La -
factored as
0.1
H, and A
(*+50.5)(j*
Thus the
1725)
is
(6-153)
l.O
1
1.4
/~\
/L
\
1.2
"
=0A
-
^=0.01^.^^
1.0
v/
0.8
0.6
0.4
0.2
i
i
0.1
0.2
0.3
0.4
0.6
0.5
Time (second)
Fig. 6-21.
A =
200.
s-plane
A
A=
A=
<**-
= 348
13.5
/41.5
A =0
A
25.25
= 13.5
2.08
-/41.5
= 348
Fig. 6-22.
is
294
Root
Sec. 6.7
295
6.7
Time Response of
The
proportional
to the
on the actuating
Qs)
Fig. 6-23.
E(s)-s(s
Analytically, Eq. (6-154)
2tco)
-i/ T,
<6
"
154 )
is equivalent to the
to the open-loop transfer function.
296
Time-Domain Analysis
of Control
Chap. 6
Systems
de(t)
dt
(c)
Fig. 6-24.
Waveforms of c(/), e(t), and de(t)ldt showing the effect of derivaStep response, (b) Error signal, (c) Time rate of change of
error signal.
The
effect
control system can be investigated by referring to Fig. 6-24. Let us assume that
is
shown
in Fig. 6-24(a).
The corresponding error signal e(t) and the time derivative of e{t) are as shown
in Fig. 6-24(b) and (c), respectively. Notice that under the assumed case the step
response exhibits a high peak overshoot. For a system that is driven by a motor
of some kind, this large overshoot is due to the excessive amount of torque
< t < t u during which the error
developed by the motor in the time interval
signal is positive. For the time interval f < t < t 3 the error signal is negative,
and the corresponding motor torque is negative. This negative torque tends to
reverse the direction of motion of the output, thus causing c(t) to undershoot
during t 3 < t < t 5 During the time interval t 3 < t < t 5 the motor torque is
again positive, thus tending to reduce the undershoot in the response caused by
the negative torque in the previous interval. Since the system is assumed to be
stable, the error amplitude is reduced with each oscillation, and the output
,
eventually
is
we can
Sec 6 7
'
297
< < is too large, and (2) the retarding torque in the time
< < is inadequate. Therefore, in order to reduce the overshoot
in the interval
interval
The
t,
/,
t2
< <
is
now
be increased in
The
cisely the
compensation
by the system of
is proportional
r d de(t)/dt. In other words, in addition to the error signal, a
signal that is proportional to the time rate of change of error
is applied to the
motor. As shown in Fig. 6-24(c), for
t
r the time derivative of e(t) is
negative; this will reduce the original torque developed due
to e(f) alone. For
< <
<t <
r,
r2
both e?(r) and de(t)/dt are negative, which means that the negative
retarding torque developed will be greater than that of the
proportional case.
Therefore, all these effects will result in a smaller overshoot. It
is easy to see that
e(/) and de(t)Jdt have opposite signs in the time interval
r2
t
t3
therefore,
the negative torque that originally contributes to the
undershoot
,
< <
is
reduced
also!
The derivative control measures the instantaneous slope of e(t), predicts the
large
overshoot ahead of time, and makes a proper correcting effort
before the overshoot actually occurs.
It is apparent that the derivative control will
affect the steady-state error of
a system only if the steady-state error varies with time. If
the
steady-state error
of a system
to de(t)jdt,
modified by
ee (s)
9e (s)
_
-
+ td s)
+ 34.5)
saq.
s(s
(6
" 155
when A
tive control.
A =
When
xd
=
= 0.1, the overshoot is completely eliminated, but the step
14 per cent.
When
%d
0, Fig.
,4
1.0
= 13.5
Td =
O.OlX^s.
//
//
0.8
^=0.1
A = 13.5
Tj=1.0
A=
0.6
0.4 ~
II
13.5
0.2
2.0
1.0
3.0
4.0
(Seconds)
Fig. 6-25. Step responses of the positional control system of Fig. 6-16 with
derivative control;
A =
13.5.
^=0.01
0.03
0.02
0.04
0.05
(Seconds)
Fig. 6-26. Step responses of the positional control system of Fig. 6-16 with
derivative control
298
A =
500.
0.08
Sec 6 7
response
is
slow in reaching
response for r d
299
1.0.
is
The
(2fa
is
*(*)
(2c0
xi(ol)s
(6_156)
+ oil
+ tco1;)s + =
2
(6-157)
Notice that the derivative control has the effect of increasing the coefficient
of
the s term by the quantity Td coJ. This means that the damping
of the system is
increased. Using the values as represented by Eq. (6-155),
the characteristic
equation becomes
s
(34.5
5Ard)s
5A
(6-158)
Figure 6-27 shows the loci of the roots of Eq. (6-158) when A
13.5 and x d is
varied from to oo. The improvement on the system damping
due to the derivative control is illustrated by the root loci of Fig. 6-28
with A set at 1500. Notice
that for small values of r, the roots of the characteristic
equation are still complex, although they are farther away from the imaginary
axis than those when
Td
0.
For
become
real.
/w
i-plane
'
Td
rd =0
-*
rd
fc
32.4
=Q
**-
Td ='
-2.1
Fig. 6-27.
It
300
Time-Domain Analysis
of Control
Systems
Chap. 6
-/86
s-plane
*d
J<
--
-/86
Td
=0
more complex
transfer func-
6.8
Time Response of
The counterpart of
derivative control
is
signal in proportion to the time integral of the error. Figure 6-29 illustrates the
The
signal
instantaneous error signal, the other, to the time integral of the error.
parameter
the system
is
is
C(s)_
E(s)
R(s)
>l
The
(6-159)
HgH
+ Kj
+ 2Ca>)
(oljs
s (s
or
s(j
+ 2?<o)
*1
[
Jo
dr
Integral contro
Fig. 6-29.
integral
com rol.
C(s)
Sec. 6.8
tem by
system by
301
the sys-
There-
improved by an order of 1.
P U S Constant the inte Sral contrl educes it to zero. In the
case of Eq
^
cm
(6-159), the system will now have a zero
steady-state error when the input is a
ramp function. However, because the system is
now of the third order it tends
to be less stable than the original
second-order
'
system. Tn fact,
ot the system
is
may
if
be unstable.
With the
comes
0.
ec {s)
0e(s)
The
characteristic equation
j3
The
5A(s + K
_
~ ~s\s + 34.5))
t
(6
" 16
is
34.5s 2
5As
5AK,
(6-161)
OO
s-plane
A =0
-
34.5
A =oo
-
34.5 +
4 ;co
s-plane
,A
=0
=
A =0
A=0
AT,
|
1
(a)
(b)
Fig. 6-30.
+ 5AK
0,
Ki
34.5.
302
Time-Domain Analysis
of Control
Chap. 6
Systems
s-plane
A =
.4=0
-*i
-34.5
1-34.5 +
AT!
T
ooll
(c)
Root
loci
K >
x
varies
34.5.S 2
5As
5AK\ =
0,
34.5.
when A
all lie
lies
and
between
34.5, the
When
K =
t
becomes second order and the two roots lie on the yea-axis for all
between and oo, and the system is on the verge of instability. When
values of A
the value of
all
is
unstable for
values of A.
As an alternative, we can fix the value of A and show the effect of varying
on the roots of the characteristic equation. Figure 6-3 1 shows the root loci
K < oo.
of Eq. (6-161) with A = 13.5 and
<
To
verify the analytical results obtained in the preceding sections, Fig. 6-32
shows the step responses of the system with A = 13.5 and K = 1, 10, and 34.5.
root loci of Figs. 6-30 and 6-31, the response for K = 34.5
x
As predicted by the
is
6.9
a pure sinusoid.
The philosophy of using the derivative of the actuating signal to improve the
damping of a system can be applied to the output signal to achieve a similar
s-plane
K,
AT,
=0
K
-32.4
1_
ATj
= M).
1.0
2.0
3.0
= 0, A =
13.5.
=34.5
4.0
5.0
Time (seconds)
Fig. 6-32. Step responses of the
positional control system of Fig 6-16
with
and integral control,
13.5.
La =
A=
303
304
Time-Domain Analysis
Chap. 6
of Control Systems
In other words, the derivative of the output signal is fed back and compared with the reference input. In practice, if the output variable is mechanical
displacement, a tachometer may be used which converts the mechanical displacement into an electrical signal that is proportional to the derivative of the
effect.
is
>, <),/\ ?)
R(s)
w2
^9
t
s(.s
*-
C(s)
+ 2?oo)
K,s
The closed-loop
<M
col
R(s)
and the
s*
characteristic equation
s
(2{a>
K,col)s
is
(6-162)
+ col
is
(2co
Kfi>l)s
col
(6-163)
Comparing Eq.
K,.
Therefore,
we conclude
it
should be noted
that the closed-loop transfer function of Eq. (6-162) does not have a zero,
and thus the responses of the two systems will not be identical even if K,
equals x d
The open-loop
is
E(s)
The system
is
is still
of type
1,
s(s
2co B
(6-164)
K,col)
not altered. However, for a unit ramp function input, the steady-state error to
Sec. 6.10
6.10
is 2Clco,
305
whereas that
n.
One of the
design techniques in
is
C()
E(s)
which
is
decomposed by
=
s(s
(6-165)
+ 2Cca)
ically accessible,
we may
C(s)
col
s*
(2Co>,
(6 " 166)
+ g )s +T>
t
CO
-O
c
2f"
(a)
-Si
(b)
a second-order system by
state feedback.
306
Time-Domain Analysis
of Control
Comparing
Chap. 6
Systems
if
we
col
Since the
eigenvalues (or the poles of the closed-loop transfer function) of a linear system
directly control the transient response of the system,
designer
is
tions. It is
it
would be
desirable
if
the
shown
Chapter 11 that
in
if
a system
is
an
gives
illustration
on how the
state
feedback and
Example
6-3
G(s)
v
'
= i^
=
E(s)
2 ,
s 2 (s
is
(6-167)
,,
1)
Figure 6-35(a) shows the state diagram of G(s), and Fig. 6-35(b) shows the state diagram with feedback from all three states. The closed-loop transfer function of the sys-
tem
is
W)
=
*
+ (g +
3
I)* 2
(6 " 168)
+ Sis + gl
we desire to have zero steady-state error with the input being a unit
and in addition, two of the closed-loop poles must be at s = 1 +j
and s = 1 j. The steady-state requirement fixes the value of gi at 20, and only
#2 and - 3 need to be determined from the eigenvalue location.
Let us assume that
step function,
The
Equating
(g 3
Ds
+ Sis +
is
-j)(s
-/)(>
+ a)
g2
and the
20
third pole
is
at s
=22
and
g3
we
(6-169)
get
11
10.
damping
ratio of 0.707,
and the
third
pole
is
left
20
o-
-O
O-
(a)
(b)
by
state feedback.
Example
6-3.
307
308
Time-Domain Analysis
of Control Systems
Chap. 6
REFERENCES
Time- Domain Analysis
1.
2.
3.
IEEE
"On Approximating
IEEE
5.
in Multi-input Controllable
Linear Sys-
PROBLEMS
6.1.
For each
(b)
(c)
(d)
6.2.
velocity,
The open-loop
given by
(a)
G(s)
50
+ 0.1.s)(l + 2s)
K
=
5(1 + 0.l5)(l + 0.5s)
K
= z
s(s +4s + 200)
(1
,
(b) G(s)
(c)
6.3.
G(s)
6.4.
in
ramp
riA
is
500
Evaluate the error series for the system. Determine the steady-state error of the
system when the following inputs are applied:
Chap. 6
Problems / 309
(a) r{t)
(b) r(t)
= ,(r)/2
= (1 + It + t*)u,0)
f
Show
In Problem
/
- 0,
for co
5 rad/sec.
What
when
r(t) is
sinusoidal?
6.6.
= 4.5 sec
7 sec
12
6.7.
6.8.
IS
A step motor gives a single step response shown in Fig. P6-8 after a pulse
Step position
Step position
0.005 sec
Figure P6-8.
exci-
310
Time-Domain Analysis
6.9.
of Control Systems
Chap. 6
Find a
tation
is
motor
applied.
model the
The attitude control of the missile shown in Fig. P5-1 1 is accomplished by thrust
The transfer function between the thrust angle 5 and the angle of
vectoring.
attack
can be represented by
(refer to
Problem
system
trol
attitude
Or
-P~
.
5.1 1)
where
The
attitude con-
is
-^c^
J
Gp (s)
Attitude rate
sensor K,s
Attitude
sensor Ks
Figure P6-9.
(a)
is
in operation
= 0).
(b)
(c)
6.10.
The
control
is
glXi
giX 2
damping
(c)
6.11.
in the gi versus
g 2 plane on which
is
= 0.707
and
co
= V/T rad/sec.
= Ax + Bw
Chap. 6
Problems / 311
where
1
B
-1
The input u
satisfies
6.12.
is
-2
Ax + B =
Repeat Problem
6.11
0.
when
0"
-2
6.13.
is
-3
represented by
u(t),
is
shown
by
d(s)
U(s)
^r\
e
Gc (s)
L
Js*
8o
L/Js 2
Desired output
-*-
Figure P6-13.
in Fig. P6-13
312
Time-Domain Analysis
of Control Systems
Chap. 6
G c (s), and
(s) is
output.
(a)
(b)
(c)
With Gc (s) = 1, determine the response of the system, 9 a (t), when the input
6 r {t) is a unit step function. Assume zero initial conditions. Discuss the
effects of L and / on this response.
Let Gc (s) = (1 + Td s), L = 10, and J = 1000. Determine the value of Td
so that the system is critically damped.
It is desired to obtain an output response with no overshoot (see Fig.
P6-13); the response time may not be minimum. Let Gc (s) = 1, and the
system is controlled through the input 6 r (t), which is chosen to be of the
form shown. Determine the values of k and f so that the desired output
is obtained. Use the same values of L and / as given before.
i
6.14.
A
V
R(s)
is
shown
in Fig. P6-14.
C(s)
Amplifiei
Motor I
gain
+o
as
Figure P6-14.
(a)
K=
For
10, determine the values of a and b to give an overshoot of 16
per cent and a time constant of 0.1 sec of the system response to a unit step
input.
Time constant
is
is
decreased
slightly,
how
does
damping factor.
damping ratio
affect the
it
of the system?
(c)
6.15.
on
The parameters of
infinity.
shown
in Fig.
below:
JL
BL =
Jm
0.00143 ft-lb/rad/sec
8 x 10"* ft-lb/rad/sec 2
= motor inertia
Bm =
Rf =
motor viscous
Lf
R
=
=
generator
total
negligible
friction
50
field resistance
La =
(a)
total
henry
motor
motor torque constant
48.8 2
K
Kg = generator constant
t
ft-lb/rad/sec 2
load inertia
0.812 ft-lb/A
200 volts/amp
motor
For an amplifier gain of K
negligible
For
K=
L (,O
when 0,(0
is
a unit step
Chap. 6
Problems
313
-A/VWr-
Rf
'f
WW*
Amplifier
gain
Lf
e
f
Ik
~T~
"
'
(g\ eg
= :onstan1
'/
+
eb
iflZ>
Motor
to load gear
ratio
Load
i
50
to potentiometer
gearratio =1/1
Figure P6-15.
(c)
(d)
Repeat parts
Repeat parts
ot
(a)
(a)
6.16.
motor
Demodulator
ed
dc amplifier
60 Hz
ac
o-
Synchros
Figure P6-16.
'f
steady-state response
shown
in Fig.
~ constant
P6-16
314
Time-Domain Analysis
of Control
Systems
Chap. 6
0.005 ft-lb/rad/sec
0.05 ft-lb/rad/sec 2
negligible
ft-lb/amp
negligible
ion
1
volt/volt
1:
(a)
Write the characteristic equation of the system and determine the value of
the dc amplifier gain A for a critically damped system.
(b)
For a
= tu {t),
value of
6.17.
volt/rad
0.05 ft-lb/rad/sec 2
In the feedback control system shown in Fig. P6-17, the sensitivity of the synchro error detector is 1 V/deg. After the entire system is set up, the transfer
function of the two-phase motor is determined experimentally as
Km
s(l
where
Km
10 volts-sec, and
Tm =
+ Tm s)
0.1 sec.
/
/
ec
=eL
,e m
2-phase
servomot
=iooei
Figure P6-17.
(a) If the
is
the
is
to be driven in
minimum
its
value of gain
of the amplifier in
order that the deviation between output and input positions will not exceed
(b)
the
damping
ratio
Chap 6
'
Problems
(c)
The
amplifier
is
of the amplifier
315
written
e 2 {t)
Ae(t)
+ ATd d-^-
A
35. Determine the value of Td so that the damping ratio
per cent. Repeat part (a) with the modified
amplifier.
where
is
40
7
Stability of Control
7.1
Systems
Introduction
It
was shown in Chapter 6 that the transient response of a linear feedback control
is governed by the location of the roots of the characteristic equation.
system
may be
regarded as
a way that the corresponding system will perform according to the prescribed
specifications.
must
We shall learn in
Chapter
Among
the
1 1
316
Sec 7
'
'
Stability. Characteristic
In this chapter
we
317
is
When
infinity.
2.
3.
If
If
sinusoidal oscillations.
sponses.
7.2
and u
state vector
homogeneous
= Ax(t) + Bu(r)
state
(7
.jx
as the
tzz-jzsr
as
is
ms
is
),
returns to the
said to bistable-
* k
*"
//^\
near Ume-mvanant
there
system
is
is
\\x(t)\[<M
for all
and
(2)
lira
t-">3
||
x(r)
||
t>t
rj_ 2
(7-3)
318
Stability of Control
Chap. 7
Systems
where
x(t)
represents the
norm* of the
l*(OII
11/
(7-4)
_(=1
any
by the norm of the vector x(r) must be bounded. Equation
that the system must reach its equilibrium state as time approaches
The condition
(>
=r
as represented
(7-3) states
infinity.
The
is
illustrated
by the second-order case shown in Fig. 7-1. The state trajectory represents the
transition of x(t) for t > t from a finite initial state x(r ). As shown in the figure,
x(t ) is represented by a point that is the tip of the vector obtained from the
forms the upper bound
vector sum *i(r ) and x 2 (t ). A cylinder with radius
for the trajectory points for all t > t and as t approaches infinity, the system
reaches
its
0.
*2<>o)
Next we
shall
show
of the characteristic equation. For the zero-input condition, the state transition
equation of the system is
x(f)
<J>(r
(7-5)
(.W,)
The norm
number.
of a vector
is
(7-6)
IIW-'o)x('o)l
||x|| is
always a real
Sec. 7.3
An important property
of the
norm of a
vector
319
is
l|x(0||<||<K'-'o)l|||x(f
which
(7-7)
)|[
- OH
lim
||
>
Mt-t) =
(7. 8 )
||
or
! *'& -
')
(7-9)
n, where
1, 2,
(t
,
/) is the yth element of
<f>u
In Eq. (4-42) the state transition matrix is written
i,j
<(>(0
<f>(t
- /).
- [(iI-A)-']
1
(7 .10)
Since si
A|
is the characteristic equation
of the system, Eq (7-11)
implies that the time response of
ft/) is governed by the roots of the characteristic equation. Thus the condition
in Eq. (7-8) requires that the roots
of the
|
characteristic equation
7.3
Although the
must
all
have negative
real parts.
preceding section
is
systems
system
is
is
as follows
K0l<JV<oo
for/>/
(7 . 12)
then
c(t)
I
< M < oo
for
>
(7_ 13 )
A differentiator
known
is
to
have
a'n infinite
We
all
320
Stability of Control
Chap. 7
Systems
Let us express the input-output relation of a linear system by the convolution integral
= T r{t - z)g(r) dr
c(t)
(7-14)
(7-14),
we
get
= \r r{t-x)g{x)dx
I
I""
r(f
is
\c(t)\<
(7-15)
- t)1
is
written
*<t)
(7-16)
rfr
Now
if r{i) is
Therefore,
if c{t) is
< Jo" N
f
to be a
g(z) dx
|
= N f"
JO
g(r) dx
(7-17)
bounded output,
[\g{T)\dT<M<oo
(7-18)
or
r|g(T)|^T<
P<oo
(7-19)
J o
(7-19)
is
evaluated from
g(t),
to
oo,
must be
finite.
We
shall now show that the requirement on the impulse response for
can be linked to the restrictions on the characteristic equation roots.
By definition, the transfer function G{s) of the system and the impulse response
g{i) are related through the Laplace transform integral
stability
G(s)=
f g(t)e'"dt
(7-20)
left
\G(s)\<r\g(t)\\e-"\dt
(7-21)
-I
The
roots of the characteristic equation are the poles of G(s), and when s
oo. Also, s
a +jco; the absolute value of
e~"
is
e~"
\.
(7-22)
one or more roots of the characteristic equation are in the right half or
0, and thus \e~'"\<N= 1. Thus
on the imaginary axis of the s-plane, a
If
>
Sec. 7.4
Methods
Eq. (7-22)
is
321
written
<f N\g(t)\dt = j\ g
(t)\dt
(7-23)
for Re(s)
Systems
> 0.
]<x>
v
Stable
region
Stable
region
^\ /
/
^
l
Oy
K
'*
s-plane
Unstable
region
Unstable
region
\/
Fig. 7-2. Stable
7.4
Methods of Determining
Although the
stability
investigating the
in the .s-plane.
Systems
may
be checked by
by finding
the roots of the characteristic equation,
these criteria are difficult to implement
in practice. For instance, the impulse
response is obtained by taking the inverse
Laplace transform of the transfer function,
which is not always a simple taska similar process is required to evaluate the
state transition matrix Mt)
The
solving of the roots of a high-order
polynomial can only be carried out' by a
digital computer. In practice, therefore,
the stability analysis of a linear system
322
Stability of Control
is
Systems
Chap. 7
seldom carried out by working with the impulse response or the state transiby finding the exact location of the roots of the characteristic
equation. In general,
we
1.
Routh-Hurwitz criterion an algebraic method that provides information on the absolute stability of a linear time-invariant system.
The criterion tests whether any roots of the characteristic equation
1
The number of
lie
and
roots that
lie
on
cated.
2.
Nyquist criterion 6
on the
difference
we know
When
closed-loop system
4.
8):
is
when
roots
lie
unstable.
function G(s)H(s)
5.
7.5
Routh-Hurwitz Criterion
1"
The Routh-Hurwitz
criterion represents a
location
zeros.
single input
and output
single- or multiple-loop
systems.
of the form
F(s)
where
all
s"
a,5"-'
a z s"' 2
an _
an
(7-24)
Sec. 7.5
it is
Routh-Hurwitz
Criterion /
2.
with
it is
How-
still
halfofthej-plane.
323
D =
t
a3
D2 =
ai
a2
a,
a3
do
a2
lie in
Dk
*3
a\
o3
a5
2-l
a2
a4
2-2
a,
a3
02-3
a2
2-4
"t
where the
(7-25)
2-5
replaced by zeros.
mtoa
in
e/S
C baSfC
kWS f a
~=
'
gebra
'
all
work with
Zl
=(-!)"
(73
the ratios
p&rt.
must be
positive
simplified
by Routh
obse
roots
AH
was
the rule
products of
all
at a time
taken 3 at a time
roots
at least
real
324
Systems
Stability of Control
The
Chap. 7
first
Routh-Hurwitz criterion is to
first row consists of the
coefficients, and the second row consists of the second,
coefficients, as shown in the following tabulation
first,
The next
step
is
ao
55
a\
a\a 2
s*
a\
a&
as
...
a-$
as
an
09
...
shown
is
Aai
s2
a az
a\a*
BC - AD
c
ED Cat,
E
Fa 6 - E x
E_
Aai
a6
<H
a5
aoa;
=B
aia *
a\a6
a\
a\B
ai
a3
a\
S3
az
a\
s"
ao
A
Ca 6 - A x
a\
xO -
= a6
x0 _
i-
CxO-/ixO =
06
F
r.
a6
The array of numbers and operations given above is known as the Routh
Routh array. The column of ^s on the left side is used for
tabulation or the
identification purpose.
Routh tabulation has been completed, the last step in the Routhis to investigate the signs of the numbers in the first column
tabulation. The following conclusions are drawn The roots of the poly-
Once
the
Hurwitz criterion
of the
nomial are all in the left half of the s-plane if all the elements of the first column
of the Routh tabulation are of the same sign. If there are changes of signs in the
elements of the first column, the number of sign changes indicates the number of
:
The reason for the foregoing conclusion is simple, based on the requirements on the Hurwitz determinants. The relations between the elements in the
first column of the Routh tabulation and the Hurwitz determinants are:
a,
=a
=D
E-S
D
'-a
Sec 7 5
Routh-Hurwitz
'
Criterion /
325
Example
7-1
- 2)(s +
1)0
= s - 4s 2 + s +
3
3)
(7-26)
which has two negative coefficients. Thus, from the necessary condition,
we know
without applying the Routh-Hurwitz test that the equation has at least
one root with
positive real parts. But for the purpose of illustrating the
Routh-Hurwitz criterion, the
Routh tabulation is formed as follows:
s3
*2
-4
Sign change
Sign change
(-4)W-(6)(1)
(2-5X6)
-(-4)(0)
=25
= 3.
Example
7-2
+ s + 3s + 5s +
2
10
(7-27)
same
satisfies
Sign change
)(
Sign change
2)(5)
(-7)(5)_-dX10)
=-7
=
10
10
643
10
first
Special Cases
may
occur
326
Stability of Control
Systems
Chap. 7
occasionally
when
The
1.
Routh
first
test:
is
zero,
The elements
2.
In the
row
where a
is
Example
7-3
will all
may be
situation
1)
determine
all zero.
first
in the next
in
term
one root of the equation
how many
= s* -
2)
is
3s
we know from
zero,
(7-28)
tabulation as follows
.$3
-3
s2
2
o
Sl
row
the factor
one's
is infinite.
(s
mind
To
first
where a is
However, for reasons that will become apparent
or 2. Let a = 3; then Eq. (7-28) becomes
a),
is 1.
choose a to be
(s
The Routh
first
l)\s
element of the
we
+ 2)(s +
3)
- si +
3s 3
-3s 2 -7s +
later,
=0
we do not
(7-29)
is
-3
-7
Sign change
-?
+ 7_
~
Sign change
(- jX-7)
18
2Q
Since there are two changes in sign in the first column of the Routh tabulation, the
equation has two roots in the right half of the f-plane.
As an
alternative to the
*If
one chooses to use a negative number for a, the (s + a) term will contribute a root in
and this root must be taken into account when interpreting the
Routh
tabulation.
.:
Sec 7 5
-
Routh-Hurwitz
number
Criterion
327
we may
Sl
-3
Sign change
~3e -
Sign change
s
On
other hand,
we may assume
e to be negative,
and we can
the
may
exist
2.
3.
1.
2.
3.
Take the
formed
tabulation.
Example
7-4
number
that
comes
into one's
l) 2 (s
2)(s
1)
(s
mind would be a
1), we have
= j* + ^ _
3^2
_^+
I.
Multiplying
(7 . 30)
328
Stability of Control
Systems
The Routh
Chap
tabulation
made
is
as follows:
s4
s3
^V
-10
-3
ax 3) ax
--2,
^_2=0
row contains
Since the s 1
has a root
(2).
at s
To remedy
1,
Routh
the
all zeros,
by the factor
(s
is,
the
row
A(s)
-2s 2
dA(s)
ds
the
row of zeros
The
(7-32),
terminates prematurely.
1).
we form
this situation,
Now,
test
in the
Routh
+2=0
(7-31)
s gives
-4.s
tabulation
(7-32)
is
s>
s2
i1
-2
-4
-3
-1
Sign change
2
[coefficients of dA(s)jds]
Sign change
Since the preceding tabulation has two sign changes, the equation of Eq. (7-28)
has two roots in the right-half plane. By solving the roots of the auxiliary equation in
Eq. (7-31), we have
s1
or
= 1
These are also the roots of the equation in Eq. (7-30). It should be remembered
that the roots of the auxiliary equation are also roots of the original equation, which is
under the Routh test.
Example
7-5
In this example we
Consider
(s
is
known
which
to have
1
1
1
As
signs at i
is
-2
-3
-3
-7
-4
-4
-4
(7-33)
= 2
and
Sec 7
Routh-Hurwitz Criterion
Since a
329
row of
coefficients
= s* -
A(s)
The
3s 2
(7-34)
is
-jf = 4s -6s =
3
(7-35)
from which the coefficients 4 and -6 are used to replace the row of zeros
tabulation. The new Routh tabulation is
-2
-3
-3
-6
-1.5
-4
Routh
in the
-7
-4
-4
[coefficients of
Sign change
^^]
16.7
-4
Since there is one change in sign in the first column of the
new Routh tabulation,
Eq. (7-33) has one root in the right half of the j-plane. The
equal, but opposite roots
that caused the all-zero row to occur are solved from
the auxiliary equation. From
Eq. (7-34) these roots are found to be
1
s3
The Routh-Hurwitz
34.5s 2
criterion
7500/^,
(7-36)
may
is
j2
,.,
7500s
7500
34.5
258,750
7500ATi
7500A-,
341
s
7500ATi
258,750
7500/sT,
343
7500*,
From
>
>
(7-37)
7_ 38)
we have
AT,
< 34.5
(7.39)
330
Systems
Stability of Control
Cn ap 7
>
Ki
Therefore, the condition for stability
is
<K <
x
Example
(7-40)
(7-41)
7-6
system,
s3
It is
3Ks 2
(K
2)s
+4=
From
(7-42)
is
stable.
The Routh
tabula-
is
s3
s2
3K
so
(K
2)
is
K>
and from the
3K + 6K - 4 >
2
When
or
K<
-2.528
or
K>
and
K>
more stringent
one.
Thus
the conditions of
K>
is
it
0.528
is
K must satisfy
K>
The requirement of K
It
<
2.528
is
0.528
disregarded since
K cannot be negative.
is
algebraic and
all
is
valid only
if
Another limitation of the Routh-Hurwitz criterion is that it offers information only on the absolute stability of the system. If a system is found to be
stable by the Routh's test, one still does not know how good the system is
in other words,
how
shown
to be unstable, the
Routh-Hurwitz
On
criterion gives
if
the system
is
no indication of how
7.6
Thus
far,
stability
The
Sec. 7.6
Nyquist Criterion
on a
digital
331
finding subroutine.
The
2.
imaginary
criterion.
These two methods are not very useful for design purposes.
It is necessary
to devise stability criteria that will enable the
analyst to determine the relative
stability of the system. It would be desirable
if the criterion indicates how the
stability of the system might be improved.
2.
3.
it
desirable
the system.
4. It
5.
make
It
We may
is,
= Ax(0 + Be(t)
c{t) = Dx(r)
- c(t)
e(t) =
x(?)
(7.43)
(7.44)
/(/)
(7.45)
When
the feedback
is
= H(sl ~ A) "'I*
(7 . 46)
>
G(s)
R(s)
"V*?
Fig. 7-3.
back.
(si--
A) IB
C(s)
332
Systems
Stability of Control
Chap. 7
*~ C(s)
Fig. 7-4.
back.
The closed-loop
CO)
7-4.
is
G(s)
R(s)
(7-47)
G(s)H(s)
byFO). Then
F(s)
= +
G(s)H(s)
= + D(sl - A)-'B#(s)
(7-48)
easy to see that the zeros of F(s) are the roots of the characteristic equation
of the system, if there is no pole-zero cancellation in G(s)H(s).
It is
Let us assume that G(s) and H(s) are rational functions; that is, G(s)H(s)
a quotient of two polynomials with constant coefficients. In general, F(s) can
be written
is
+ z,)(j + z ) (2 + z m )
(7-49)
+ pMs + p )...(s + p
where z,(i = 1,2, ... ,m) &ndp k (k =
n) are either real or in complex2,
conjugate pairs. Then the roots of the characteristic equations are s = z
F(s)
K(s
s J (s
1,
n)
t ,
z 2
zm
to be stable,
it is
of these roots has a positive real part. There is no particular restriction on the
location of the poles of F(s), which are at s
, p.
0,
It is
p l7 p z
important to note that the poles of F(s) are the same as those of G(s)H(s). If
lies in
still be
of F(s) are found in the left half of the ,?-plane. This is a
very important feature of a feedback control system. In preceding chapters it
has been pointed out that a high forward-path gain generally reduces the steady-
is
state error
it is
may
result in
made
an
stable
by proper design.
Before embarking on the fundamentals of the Nyquist criterion, it is essential to summarize on the pole-zero relationships with respect to the system
functions.
Sec. 7.6
Nyquist Criterion
1.
Identification of poles
2.
The
=
=
=
333
and zeros:
= zeros of G(s)H(s)
= poles of G(s)H(s)
poles of C(s)/R(s)
zeros of F(s)
= +
1
G(s)H(s)
transfer
function G(s)H(s).
3.
located in the
left half-
of the
.s-plane.
Encircled.
the path.
since
the application of the Nyquist criterion, the closed path also has a direction
associated with
it.
As shown
is
said to be encircled
by
When
considering
all
we can
say that
334
Systems
Stability of Control
Chap. 7
Enclosed.
point or region
is
(b)
(a)
by T.
(b) Point
Number of
is
and
encirclement
to
enclosure.
number
When
N may be
is
enclosed
by the locus T.
point
is
assigned to the
encircled or
number of
en-
and
B is enclosed
may
be.
then
let s t
(a)
(b)
enclosure.
Sec. 7.6
encircled twice,
once; point
all in
is
Principle of the
enclosed twice.
Argument
is
Suppose that a continuous closed path r, is arbitrarily chosen in the splane, as shown in Fig. 7-8(a). If all the points on I\ are in the specified region
,,
ly'ImF
](X>
s-plane
F(s)-plane
*-
(a)
(b)
ing locus
which F(s)
in
F(s)-plane
point s
locus
is
ReF
is
is
TF
in the
Correspond-
FO)-plane.
F(s
) is
and then returning to jj after going through all the points on the
shown in Fig. 7-8(a)], the corresponding F P locus will start from
point F(sj) and go through points F(s 2 ) and F(s 3 ), which correspond to s 2 and
s 3 respectively, and return to the starting point, F(Ji). The direction of traverse
of r F may be either clockwise or counterclockwise; that is, in the same direction
or the opposite direction as that of r s depending on the particular function
(clockwise)
locus [as
F(s).
rP
is
shown, for
illustration purposes, to
be
counterclockwise.
It
should be pointed out that, although the mapping from the j-plane to
is one to one for a rational function F(s), the reverse process is
the F(j)-plane
336
Stability of Control
Systems
Chap. 7
the .s-plane.
maps
s{s+l)(s
The
left side
F(s)
is
of Eq. (7-51)
is
Eq. (7-50) as
+ 2)=-^
(7-51)
chosen to be a constant.
The
principle of the
is
is
argument can be
that
s;
number of
the zeros
the
and
the
is
chosen
in the s-plane
many
mapped
by
IV
N= Z - P
(7-52)
where
N=
Z=
origin
made by
the F(s)-plane
rF
in the
in the
s-plane
encircled
by the
.s-plane locus
.s-plane
N>
> P).
(Z
= P),
or negative (Z
< P).
is
as that of 1%.
2.
/V
(Z
TV
<
(Z
< P).
direction
FF
is
from that of T
A convenient way of determining N with respect to the origin (or any other
point) of the F(s) plane
is
to
draw a
line
in
any direction to
Nyquist Criterion
Sec. 7.6
F(s)-plane
337
F(s)-p\ane
N=-2
F(s )-plane
N=
Fig. 7-9.
infinity; the
number of net
N in the F(i)-plane.
may be considered as a heuristic explanation of the prinLet us consider the function F(s) given by
following illustration
ciple.
F(s)
where
K is a positive
(7-53)
+ />i)C* + Pi)
written
\F(s)\/F(s)
+ Pi
+p
Us
2
Is
+ Pi js + p
(7-54)
2)
338
Stability of Control
Systems
Chap. 7
x-plane
ReF
of
(a)
(7-53).
5^
at s
st
given by
K(s
F( Sl )
0*1
zt)
(7-55)
+ />i)C*i + Pi)
it
returns to the
Sec. 7.6
if there were any) that are not encircled by T s when Si completes one
round trip are zero; whereas the vector (j, + z,) drawn from the zero at z u
which is encircled by I\, generates a positive angle (counterclockwise sense)
of 2n rad. Then, in Eq. (7-54), the net angle or argument of F(s) as the point
Si travels around T s once is equal to 2n, which means that the corresponding
F(s) plot must go around the origin 2n radians or one revolution in a counterclockwise direction, as shown in Fig. 7-10(b). This is why only the poles and
zeros of F(s), which are inside the JT, path in the j-plane, would contribute to
the value of N of Eq. (7-52). Since poles of F(s) correspond to negative phase
angles and zeros correspond to positive phase angles, the value of
depends
only on the difference between Z and P.
zeros
Z=
F=0
Thus
N = Z- P=
which means that the F(y)-plane locus should encircle the origin once in the same
direction as the s-plane locus. It should be kept in mind that Z and P refer only
to the zeros and poles, respectively, of F(s) that are encircled by r and not the
total number of zeros and poles of F(s).
In general, if there are
more zeros than poles of F(s), which are encircled
by the j-plane locus r, in a prescribed direction, the net angle traversed by the
.F(j)-plane locus as the s-plane locus is traversed once is equal to
2n(Z
-P) = 2nN
(7-56)
This equation implies that the F(s)-plane locus will encircle the origin
N times
same direction as that of r,. Conversely, if N more poles than zeros are
encircled by T s in a given prescribed direction, N in Eq. (7-56) will be negative,
and the F(s)-p\ane locus must encircle the origin times in the opposite direction
in the
to that of
A
is
IV
summary of
all
Table 7-1
Summary
Principle of the
Argument
F(s)-Plane Locus
N=ZP
N>
N <0
Sense of the
Number of Encirclements
s-Plane Locus
of the Origin
Clockwise
Counterclockwise
Clockwise
Clockwise
Counterclockwise
Clockwise
Counterclockwise
Af
Clockwise
Counterclockwise
Direction of
Encirclement
Counterclockwise
No encirclement
No encirclement
340
Stability of Control
Systems
Chap. 7
Nyquist Path
At
many
if
Of
is
taken to be one
course, as an alternative,
singularity of F{s),
s-plane
Poles of
F(s)
It is
apparent that
Sec. 7.6
if
lies inside
will
it
be
For the convenience of analysis, the Nyquist path is divided into a miniof three sections. The exact number of sections depends upon how many
of those small semicircles are necessary on the imaginary axis. For the situation
illustrated in Fig. 7-1 1, a total of eight sections needs to be denned. The order
+
of numbering these sections is entirely arbitrary. The notations j co\, j co~i, +j0
+
j0 ja>U and jcoj are used to identify the starting and ending points of
mum
Section 1 from s
+j to
jcot along they'co axis.
Section 2: from +jcot to +jcoT along the small semicircle around
:
section 3
Section 4:
Section 5
=j bi-
Section 6: from j
3).
to
jco~[
j cot along
image of section
s = joo
1).
Section 8: from
to
+jo along
the
semicircle
of
infinite radius.
is specified,
the F(s)
and
G(s)H(s) locus
when
j takes
investigating the behavior of the F(s) plot with respect to the origin of the
is
function G(s)H(s)
plane. This
is
is
given, a simpler
l,y'0)
is
if
two types of
stability
We
are inter-
stability of the
that G(s)H(s) has poles only in the left half of the j-plane.
342
Stability of Control
Systems
With
two
sets
Chap. 7
this
stability
problem,
necessary to define
it is
of N, Z, and P, as follows
G(s)H(s)
P = number
N_
Z_
=
=
P_
number of zeros of 1
G(s)H(s) that are encircled by the
Nyquist path, or in the right half of the j-plane
Nyquist path, or in the right half of the j-plane
Several facts
become
clear
Po
since G(s)H(s)
and
at this point
P-i
(7-57)
Z_!
(7-58)
P =
The crux of
the matter
is
(7-59)
determined by the
is
principle of the
argument for
stability studies
summarized as follows
1.
is
is
defined according
( \,j0)
4.
point, respectively.
(if it is
not already
N =Z -P
if
Z_
is
is
given.
Once
is
it
=P
[Eq. (7-57)],
and
determined from
N. =
Since
determined, P_
(7-60)
Z_,
/>_!
(7-61)
must
-P_,
(7-62)
Sec. 7.6
may
loop system to be
point as
stable, the
many times as
the s-plane,
and
In general,
it is
fact, a great
problems can be solved simply by sketching the part of G(s)H(s) that corresponds
to the +yco-axis of the s-plane. The condition when this simplified procedure is
possible
of the
is
when P
.y-plane.
0; that
is,
JV-i
Z_,
(7-63)
which means that the Nyquist plot of G(s)H(s) can encircle the ( l,j0) point
only in the counterclockwise direction, since N_ in Eq. (7-63) can only be zero
i
is
stability
is
equivalent to enclosure,
we
critical
whether N.
only the
is
zero,
.y-plane.
The Nyquist
in the right
P_
case may be stated: If the function
half of the s-plane, for the closed-loop system to
be stable, the Nyquist plot of G(s)H(s) must not enclose the critical point ( 1J0).
Furthermore, if we are not interested in the number of roots of the char-
equation that are in the right-half plane, but only the closed-loop
only the G(s)H(s) plot that corresponds to the positive imaginary axis
of the j-plane is necessary. Figure 7-12 illustrates how the s
/oo to s
acteristic
stability,
Im
GH
G(s)#(s>plane
ReGH
Fig. 7-12.
is
enclosed.
= jco to s
0,
344
Stability of Control
Systems
Chap. 7
7.7
( l.y'O)
is
may
critical
point
enclosed.
Example
7-7
W = rh)
where
corresponds to j
it is
=/co
It is
Pa =
(7 - 64)
P_i
= 0.
To determine
to s
see
if it
encloses the
we
1,7*0)
shall construct
is
shown
in Fig. 7-13.
s-plane
shown
0.
it is
Example
7-7.
is
in Fig. 7-13.
may be
is
e">
(7-65)
where {
0) and 9 denote the magnitude and phase of the phasor, respectively.
As the Nyquist path is traversed from +j0 + to jO* along section 2, the phasor of
Sec. 7.7
i /
Im
GH
345
GCs)i/(s)-plane
s-plane
80 counterclockwise
rotation
*~
*-
(a)
co
ReGH
(b)
2.
Eq. (7-65) rotates in the clockwise direction through 180. Also, in going from
y'0 +
90 through
from +90
+j0 +
to
The corresponding Nyquist plot of
, 9
G(s)H(s) can be determined simply by substituting Eq. (7-65) into Eq. (7-64). Thus
to
varies
0.
G(s)H(s)
s=ee )o
Since e
0,
is
ee">(ee">
(7-66)
+ a)
simplified to
G(s)H{s)
ooe -je
(7-67)
aeeJ<>
which indicates that all points on the Nyquist plot of G(s)H(s) that correspond to
section 2 of the Nyquist path have an infinite magnitude, and the corresponding phase
is opposite that of the j-plane locus. Since the phase of the Nyquist path varies from
+90 to 90 in the clockwise direction, the minus sign in the phase relation of Eq.
(7-67) indicates that the corresponding G(s)H(s) plot should have a phase that varies
from 90 to +90 in the counterclockwise direction, as shown in Fig. 7-14(b).
In general,
plots, the
= and s = oo may be
problem the behavior of G(s)H(s)
determined from
For
lim ,
s ^o s(s
is
K
+ a)
.
i-
lim
s.
As
magnitude, from
+/0 + to
is
(7-68)
is
inversely propor-
346
Chap. 7
Systems
Stability of Control
lim Jfo"
(7-69)
by a phasor of
direction
if
is
counterclockwise direction
if
is
The
traced out
used.
is
infinitesimally small
points
on the
semicircle
5
may
infinite radius
in Fig. 7-13
is
on the Nyquist
shown in
isolated, as
Rei*
(7-70)
plane
Re/*
l
Im
GH
G(,s)i/(.s)-plane
180
counterclockwise
rotation
360 clockwise
rotation
ReG
Radius
(b)
where
oo. Substituting
(b)
Nyquist plot
4.
G(s)H(s)
-_
Re j*
K
~ RW* = Oe-i
1*
(7-71)
which implies that the behavior of the G{s)H{s) plot at infinite frequency is described by
a phasor with infinitesimally small magnitude which rotates around the origin 2 x 1 80
360 in the clockwise direction. Thus the G(s)H(s) plot that corresponds to section
Sec. 7.7
jco(jco
is
rationalized
347
becomes
G(jOi)H(jOi)
which
(7-72)
+ a)
G(jco)HUco)
The
intersect of
G(jco)HQco) on the
real axis is
found from
=
co*
plane
(7-73)
Im G(jCO)H(jco)
which gives
_ K(co 2 jaco)
=
co* + a 2 co 2
co
co. This
means that
-Kaco
+ a 2 C0 2
-Ka
+a
co(cb 2
(7-74)
2
)
on the real
i /
Im
GH
G(s)H(s)-p\ane
ReGH
Fig. 7-16.
K/ls(s
a)].
348
Systems
Stability of Control
Chap. 7
and
4,
finite
axis.
It is of interest to check all the pertinent data that can be obtained from the
= iV-i = 0. By inspection of Eq. (7-64), Z =
Nyquist plot of Fig. 7-16. First,
andP = 0, which satisfy Eq. (7-60). SinceP = P-u Eq. (7-61) leads to
Z_!
Therefore, the closed-loop system
N-i
stable.
is
=0
+P_!
(7-75)
as
is
simply
+K=
(7-76)
whose roots will always lie in the left half of the i-plane for positive a and K.
Figure 7-16 also shows that for this problem it is necessary only to sketch the
portion of G(s)H(s) that corresponds to section
that the
( 1, y'O)
It is
apparent
all
positive
values of K.
Example
7-8
Consider that a control system with single feedback loop has the loop
transfer function
<*>" =
The
is
Section 2.
(1
(7 " 77)
WTTY
is
+ K)s - K =
(7-78)
all
positive K.
e ie
lim G(s)H(s)
= lim
ooe-^< + )
(7-79)
This means that the Nyquist plot of G{s)H{s) which corresponds to section 2 of the
Nyquist path is traced by a phasor with infinite magnitude. This phasor starts at an
+90 and ends at 90 and goes around the origin of the G(s)H(s)-p\ane
counterclockwise a total of 180.
angle of
Section
4.
Re'*
lim G{s)H{s)
= lim =
0e~'*
(7-80)
the Nyquist plot of G(s)H(s) corresponding to section 4 of the Nyquist path goes
around the origin 180 in the clockwise direction with zero magnitude.
Section 1. s =jCO:
Thus
G(j<d)H(jco)
JooUm
o4
1}
+W
-A
a>(co*
1)
(7-81)
at
= 1
we have
rad/sec
GUl)H(jl)
Based on the information gathered
(7-82)
real axis.
=K
Then
(7-83)
complete Nyquist
Sec. 7.7
lm
GH
G(s)H(s)
AL, =
349
plane
Fig. 7-17.
plot of G(s)H(s)
is
shown
in Fig. 7-17.
K(s
ReGH
l)/[s(s
1)].
Z =1
P =p_,
Figure 7-17 indicates that Ao
Figure 7-17 also gives A/_j =
=
1.
1,
which
is
=o
in
criterion.
Then
Z_, =AT_,
+P_, =
( 7 _84)
which means that the closed-loop system is unstable, since the characteristic equation
has one root in the right half of the 5-plane. The Nyquist plot of Fig. 7-17 further
indicates that the system cannot be stabilized by changing only the value of A'.
Example
7-9
Consider the control system shown in Fig. 7-18. It is desired to determine the range of
for which the system is stable. The open-loop
C(s)
Grrt
is
10*(j
2)
(7-85)
on the jco
axis, the
Nyquist path
350
Stability of Control
R(s)
Chap. 7
Systems
-N
Y^
E( S )
K(s + 2)
.S
Fig. 7-18.
C
C()
10
2
(.5
+ 3)
>
7-9.
.s-plane
Fig. 7-19.
in Fig. 7-19.
(7-85).
The construction of
the
Re 1 *
lim G(s)
10K
0e -j2 *
(7-86)
the phasor for section 4 of the Nyquist path is traversed from 90 to +90
through 180 counterclockwise, Eq. (7-86) indicates that the corresponding Nyquist
plot of G(s) is traced by a phasor of practically zero length from +180 to 180
As
through a
Section
1.
=jco:
l0K(ja>
2)
GUco) -(l0-3co*)-jcoi
(7-87)
G(jco)
10 AT[2(10
3CQ 2 )
-co* +7CQ(10
(10
3co 2 ) 2
co 6
3co 2 ) +./2CQ 3 ]
(7-88)
Sec. 7.7
351
co
= VlO rad/sec
rad/sec
and
which correspond to the frequencies at the intersects on the real axis of the G(s)-plane.
In this case it is necessary to determine the intersect of the G(s) plot on the imaginary
axis. Setting the real part of G(jG>) to zero in Eq. (7-88), we have
co 4
6co 2
co
1.54 rad/sec
20
(7-89)
which gives
Therefore, the intersects on the real axis of the G(.s)-plane are at
C(/0)
2 AT
and
C(yVlO)
and the
intersect
on the imaginary
axis
= -K
is
G(j^/T) =j]0/TK/3
With the information gathered
of Eq. (7-85)
is
sketched as shown
in the
in Fig. 7-20.
may
../ImC
G(s)-plane
Fig. 7-20.
\QK(s
2)/(s 3
3s 2
10).
352
Stability of Control
Chap. 7
Systems
means
that
P =
2.
Thus P_
=2. Now,
JV_,
right half
of the
A = 2.
.s-plane,
= Z_ - P_t = Z_, -
Since Eq.
Z = 0;
this
we have
(7-90)
Thus for the closed-loop system to be stable, Z_i = 0, which requires that JV_ = 2.
With reference to Fig. 7-20, the stability criterion requires that the ( 1,;0) point must
t
K>
The reader can
critical
point should
(7-91)
criterion
It
is
much
simpler to use in stability problems such as the one stated in this illustrative
7.8
informative to illustrate
it
added to a typical loop transfer function G{s)H{s). This investigation will also
be helpful to gain further insight on the quick sketch of the Nyquist locus of a
are
given function.
G(s)H(s)
for
T1
(7-92)
Ti
< ca <
a semicircle, as
is
in Fig. 7-21.
ImGH
G(s)H(s)-p\ane
CO
co
ReGH
Ik
Fig. 7-21.
Kj(\
+ T lS
).
shown
Sec. 7.8
Addition of poles at s
0.
is
353
added to the
G(s)H(s)
(7-93)
and
a,
oo, as
shown
is
is
reduced by
rotated by
90 from
GH
Im
G(s)H(s)-phne
*-ReGH
Fig. 7-22.
K/[s(l
Tis)].
magnitude of G(jco)H(j(o) at
lim IG(j<B)H(jco)
-(j
1)
(7-94)
= -j*
(7-95)
lim|GO'a>)#C/e>)|
(7-96)
\im\G{jm)H{j(o)\
(7-97)
lim /GUco)H(jco)
co~.0
G(s)H(s)
(7-98)
and
G(s)H{s)
In view of these illustrations
will affect the stability adversely,
at s
it is
s\\
K
+T
s)
(7-99)
354
Stability of Control
Chap. 7
Systems
G(s)#Cs)-plane
ReGH
K/[sHl
= K/[s 2 (l + TlS
)].
G(s)H(s)
(2)
Tis)].
When
a pole at s
l/T2
is
G{s)H(s)
+
of G(s)H(s) at co =
(1
+T
7V)(1
is
(7-100)
2 s)
pole, since
lim GUco)H(joi)
at co
lim
-K
~"
=
m ->o 1 \1
<-.a
(7-101)
is
lim G(jco)HUco)
=K
Im
GH
/-180
G(s)H(s)-p\ane
*-
G(s)H(s)
#/[(!
Tis)(l
(7-102)
2.0)
+ T2 s)(\ + T
Kj[{\
3 s)].
Re Gtf
+ TlS)(l + T2 s)].
(2)
Sec. 7.8
Thus the
Eq. (7-92)
is
effect
of adding a pole at
= \/T
to shift the
quency, as shown in
G(s)H(s)
(1
K
+ 7(1 + T
7V)(1
355
fre-
(7-103)
3 s)
These examples show the adverse effects on stability that result from the
addition of poles to the loop transfer function.
Addition of zeros.
derivative control
stable.
by 90
+ 7 to
= oo.
at co
is
given
by
^>^
,( i
+ ri(i + r
(7
2 ,)
104>
<K
T,
+ 7
+ TV)
K{\
G(s)H(s)
s(l
(7-105)
T,5)(l
The Nyquist loci of the two transfer functions of Eqs. (7-104) and (7-105)
shown in Fig. 7-25. The effect of the zero in Eq. (7-105) is to add
are sketched as
,
.
Im
GH
G(s)H(s)-p\ane
T T2
K{T + T2 )
t
-*-
KeGH
-<(&.- T)
(2)
co
Fig. 7-25.
G(s)H(.s)
Nyquist
K(l
+ Tds)/[s(l +
G(s)H(s)
TisXl
(2)
356
Stability of Control
Systems
Chap. 7
-K(T + T^IT^il + Td
t
),
which
is
= 0.
+ T )/T T to
is
plane.
7.9
Stability of Multiloop
Systems
The
how
is
is
known
We
shall
now
illus-
back loops.
In principle, all feedback systems with single input and output can be
reduced to the basic system configuration of Fig. 7-4. Then it would seem that
the Nyquist criterion can be applied in a straightforward
manner
to the equiva-
lent
to the
block
is
it is
= - Gi(j)G(j)
+ G (s)H(s)
K(s + 2)
+ 10)[j(j + 1)0 + 2) +
G(s)
(s
The
stability
(7-106)
5]
H(s) = 5
Fig. 7-26.
Stability of Multiloop
Sec. 7.9
Systems
locus of G(s), except that the poles of G(s) are not entirely known.
we can
357
To avoid
two stages, as there are two feedback loops. First, consider only the inner
whose loop transfer function is G 2 (s)H(s). We shall first sketch the Nyquist
locus of G z (s)H(s) for
co < oo. The property of the G 2 (s)H(s) plot with
respect to the ( l,jO) point gives an indication of the number of zeros of 1 +
G 2 (s)H(s) that are in the right half of the J-plane. Having found this information,
in
loop,
<
we then proceed
co
<
<
G 2 (s)H(s)
s{s
1)(j
(7-107)
2)
j\mG 2 H
,.
G 2 (sW(s)-plane
ReG 2 //
Nyquist plot of
Fig. 7-27.
Since the
itself,
( l,y'0)
point
is
G 2 (s)H(s) =
Ms +
i)(s
2)].
is
stable
by
is
left
we
only have to investigate the crossover point of the G(s) locus with respect to the
( l.v'O)
Now,
7-26.
let
range of
K for
be
stability is
is
+2
s+
__ s
(7-108)
Gl{s)
His)
~ sis +
=5
K
IX*
(7-109)
2)
(7-110)
368
Stability of Control
Systems
Chap. 7
jlmG
G(i)-plane
*-
Fig.
[K(s
2)]/{U
10)[s(s
+-
l)(j
2)
5]}.
In this case
parameter
ReG
is
we cannot use
the
method
The open-loop
W ~_
still
is
G.(j)Ga(j)
G(S )
+G
2 (s)H(.
s
(s
Since the
unknown gain
use the Nyquist criterion
However, we may
I0)[s(s
(7-111)
l)(j
+ 2) +
5tf]
would be of no
s(s
10)(j
1)(*
2)
+s+2+
5K(s
10)
(7-112)
contain K.
we
K as
do not
We have
1
+ s(s +
may be
is
SK(s + 10)
10Xs + l)(s + 2)
of the form
+ s + 2~
+ G (s) = 0, the roots
3
(7 " 113)
of the charac-
teristic
equation
investigated
Figure 7-29 shows that the Nyquist locus of G 4 (s) intersects the real axis
( 1, y"0) point. Thus all the poles of G 3 (s) are in the left half
of the 5-plane. The Nyquist plot of G 3 (s) is sketched as shown in Fig. 7-30. Since
Sec. 7.9
Stability of Multiloop
,.
Systems
359
/Im C 4
G 4 (s )-plane
0.009
ReC,
r
\
G4 (s) =
,.
(s
2)/[s(s
10)(j
1)(*
2)].
Im G 3 (s)
C 3 0?)-plane
ReG,
*-
+s+
<K<
(s)
[5K(s
W)]J[ s (s
I0){s
l)(s
+ 2)
2].
on the
real axis
is
at
stability
10.
Kh
P{s)
KQ{s)
(7 _ 115)
is
studied by
function
<*)-3gj!
(7-116)
360
Stability of Control
Systems
Chap. 7
Thus we have
also indicated a
method of applying
The Nyquist
more convenient approach than the Routh-Hurwitz criterion,
locus represents a
The
written
|*I-A|
the Routh-Hurwitz criterion
is
On
(7-117)
We can also
=
manner outlined
we can apply
is
7.10
is
and
independent
apply
in Eqs. (7-115)
all
other
Systems with time delays and their modeling have been discussed in Section
5.12. In general, closed-loop systems with time delays in the loops will be subject
to
more
stability
Td is modeled by the
no longer have constant coefficients. Therefore, the RouthHurwitz criterion is not applicable. However, we shall show in the following
that the Nyquist criterion is readily applicable to a system with a pure time
tion of the system will
delay.
Let us consider that the loop transfer function of a feedback control sysis represented by
G(s)H(s)
^e- T G
's
(s)H, (s)
(7- 1
8)
is
The
effect
GyfJo^Hiijco) at each co
Sec. 7.10
Im
361
GH
G(s)#(s)-plane
^ReGH
T,=4
0.16
Fig. 7-31. Nyquist plots of G(s)H(s)
e- T "/[s(s
l)(.s
2)].
G(s)H(s)
e-'-'G^H^s)
=
s(s
system
is
stable
riorates as
Td =
sec.
Td
Td
when
It is
IX*
(7-119)
2)
Td is zero,
increases.
This
is
point.
Unlike the rational function case, the analytical solution of the crossover
is not trivial, since the equations
that govern the crossings are no longer algebraic. For instance, the loop transfer
The
result
is
GUcomjco)
T
<
ffi
ff i - Lt
(
/CQ(2
~ G)2)1
<
7 - 120)
362
Stability of Control
Systems
The condition
Chap. 7
for crossings
3co
which
is
sin
on the
coTd
co(2
Td
of the G(s)H(s)-p\ane
real axis
co
is
cos coTd
crossover problem
is
Bode
plot
for
all
frequencies, the
Since the time-delay term affects only the phase but not the magnitude of G(jco)
H(jco), the phase of the latter
angle of
coTd
is
is
locus intersects the negative real axis. In general, analysis and design problems
more
Bode
diagram.
If the time delay is small,
it
is
e~
T"
-Ts+
d
is,
Tjs*
(7-121)
2!
3!
Figure 7-32 shows the Nyquist plots of the transfer function of Eq. (7-119)
Im
GH
G(s)//(i)-plane
KeGH
Two-term
approximation
Fig. 7-32.
Stability of Nonlinear
Sec. 7.11
with
Td =
0.8 sec,
l-7>
=
s(s
7.11
is
Criterion /
363
and
G(s)H(s)
which
Systems Popov's
l)(s
(7-122)
2)
the result of truncating the series of Eq. (7-121) after two terms
Stability of Nonlinear
Systems Popov's
Criterion 16
When
a system
is
nonlinear,
it
and therefore there are no eigenvalues to speak of. The Routh-Hurwitz criterion
becomes useless in this case. The kind of stability that we are concerned with
here is asymptotic stability, which means that the equilibrium state x e is asymptotically stable if every motion starting at an initial state x will converge
to
x e as time approaches
Popov's
infinity.
Fig. 7-33.
shown
in Fig. 7-33.
The
nonlinearity is described
lie
in the first
as
shown
in Fig. 7-34.
The
G(s),
2.
7-34; or
(7-123)
Me)
Fig. 7-34.
Nonlinear characteristics.
1%)
N(e)
/'
/
k,
=0
-D
_z_
=0
/
/
/
/
/
/
/
(b)
7
(c) Saturation
common
/
k,
=0
zone
tems, (a) Ideal relay, (b) Relay with dead zone, (c) Saturation. Id) Saturation with
364
dead zone.
Stability of Nonlinear
Sec. 7.11
The theorem
loop system
is
Systems
Popov's Criterion
The
365
closed-
asymptotically stable if the Nyquist plot of G(jco) does not intersect or enclose the circle that is described by
is
*i
~t~
k2
*."
(7-124)
_
LK^Ki
where x and y denote the real and imaginary coordinates of the G(ja>)plane, respectively.
It
On
system
is
mean
stable,
if
if
is sufficient
but
is
not
unstable.
The
is
is
violated,
it
Fig. 7-36.
Im G
G(/oj)-plane
ReG
It is interesting to
are with
A:,
0.
observe that
( 1
y'O)
stability,
x
For
if
point,
(7-125)
line.
366
Stability of Control
Systems
Chap. 7
Example 7-10
-1
Fig. 7-37.
The
of the system
G(s)
is
=
s(s
1)0
(7-126)
2)
It can be shown that if the saturation characteristic were absent, the closed-loop
< 6. It is desired to determine the value of so
system would be stable for
that the nonlinear system is assured of stability. Notice that the Popov criterion is of
such a nature that for the saturation nonlinearity, the result of the criterion is inde-
<K
the
Re GU<o)
The frequency
at
which Re G(Jco)
d Re
G(jco)
dm
which gives
(O
is
9co2+{2
maximum
is
~~
[9co
(2
determined from
- o>
)]
2 2 2
) ]
(7-128)
= 0. Thus
is
G(j<x>)
= -f
(7-129)
stable for
K<
is
(7-127)
2 2
)
_ <a[-18 + 4(2 - co
Max. Re
which
_ (0
(7-130)
Chap. 7
References
/'
367
Im G
ReC
Popov
REFERENCES
Routh-Hurwitz Criterion
1.
E.
J.
6,
Part
II,
Macmillan
& Co.
2.
3.
G. V.
trol,
4.
V.
S. S.
IEEE
IEEE
Stability
Criterion
and
Aug. 1972.
368
Stability or Control
Systems
Chap. 7
Nyquist
J.,
Jan. 1932.
7.
C. H. Hoffman,
teristic
8.
9.
10.
I.
C. H. Hoffman,
II.
Locating the
11.
R.
Part I,"
12.
R.
W. Brockett and J. L. Willems, "Frequency Domain Stability CriteriaIEEE Trans. Automatic Control, Vol. AC-10, pp. 407-413, Oct. 1965.
Part II,"
13.
14.
IEEE
Trans.
Crite-
T. R. Natesan,
IEEE
rion,"
AC-U,
"A Supplement
Popov's Criterion
16.
V.
17.
Z. V. Rekasuis,
Element,"
18.
C. A. Desoer,
"A
Generalization of the
Popov
Criterion,"
IEEE
Trans. Auto-
S.
C. Pincura,
"On
IEEE
Popov
AC-12, pp.
Y. S. Cho and K. S. Narendia, "An Off-Axis Circle Criterion for the Stability of
Feedback Systems with a Monotonic Nonlinearity," IEEE Trans. Automatic
Control, Vol. AC-13, No. 4, pp. 413-416, Aug. 1968.
21.
J.
B.
Moore, "A
C. A. Desoer,
Control, Vol.
23.
E. Y.
"An Extension
No.
1,
IEEE
IEEE
Trans. Automatic
No.
3,
Problems
Chap. 7
369
24.
25.
26.
PROBLEMS
7.1.
By means of
the
Routh-Hurwitz
criterion,
determine the
stability
of systems
that have the following characteristic equations. In each case, determine the
(b)
(c)
(d)
(e)
7.2.
+ 20s + 9s + 100 =
s + 20s + 9s + 200 =
3s* + 10s + 5s + s + 2 =
s* + 2s + 6s 2 + 8s + 8 =
s
+ 2s + 8s 4 + 12s + 20s 2 +
(a) s 3
The
16s
16
system.
(a) s*
(b) s*
(c)
7.3.
s3
+
+
+
+ \0s 2 +2s + K =
+ 5s 2 + (10 + K)s +
(K + 0.5)s 2 + AKs + 50 =
22s 3
20Ks 3
15
The open-loop
G(S)
It is
desired that
region to the
left
left
all
given as
is
= s(l+7s)
of the line s
= a.
lie
in the
is obtained, but also that the system has a minimum amount of damping.
Extend the Routh-Hurwitz criterion to this case, and determine the values of K
and T required so that there are no roots to the right of the line s = a.
system
7.4.
The loop
^^- s(i+
rsxi
is
given by
1)
h2s)
is
stable.
The open-loop
GM =
y '
K{s
s 3 {s
+ 5X* + 40)
+ 200)(s + 1000)
is
given by
370
/ Stability of Control
Systems
Chap. 7
What
7.6.
A controlled process
is
x2
The
control
is
= Xi 3X2
= 5X\ + u
= ^1*1 + glXl
"
7.7.
Given a
is
equations
(t)
= Ax(?) + B(?)
where
B =
-2
-3
.0
-o
(T
A=
_1
implemented by
is
= -Gx(t)
G = [gi gz
u(t)
where
overall system
7.8.
gi\ with
is
to
so that the
asymptotically stable.
= Ax +
B, where
A=
on
B =
-2
3_
g u g 2 and g 3 equal
real constants.
may
_1
be implemented.
Is the
system stabilizable by
closed-loop system.
(a)
G(s)H(s)
(b) G(s)H(s)
( \
<CJ
r<\n<\
U(S)H(S)
= s{l
+ Q J)(1 + Q2s)
= j2(1
100(1
j(1
(d) G(s)H(s)
25s){l
]i)(1
+ Q5s)
+ s)
+ a5j)(1 + Q8s)
0.25s)
Chap. 7
Problems
(e)
G(s)H(s)
7.10.
G(s)H(s)
+ 02s)(s -
1)
ffi+ffi
Sketch Nyquist diagrams for the following loop transfer functions. Sketch only
the portion that is necessary to determine the stability of the closed-loop system.
Determine the
(a)
G(s)H(s)
(b) G(s)H(s)
stability
(c)
of the systems.
100
s(s 2
+2s +
2)0
1)
50
=
s(s
2)(s 2
4)
G(s)H(s)
1
7.11.
371
10
=
j(1
(f)
- 0.2s
Figure P7-1 1 shows the entire Nyquist plots of the loop gains G(s)H(s) of some
feedback control systems. It is known that in each case, the zeros of G(s)H(s)
Im
-0
oo
cj
= +
Arrows on diagram
correspond to defined
sense of the Nyquist
path
"".^
G//-plane
j-- Re
372
Stability of Control
Systems
Chap. 7
co
= -
/Im
~~ " ^^
"""
X
/
(///-plane
/?-*
W *Ju
= -=
/**'
=+
(-1./0)
co
Re
'
CO= +
(c)
are
all
located in the
left
is,
Z=
in
the right half of the s-plane. State the stability of the open-loop systems. State
is
stable ;
if
The
5Ks
+ (2K +
3)s
10
is
given by
Apply the Nyquist criterion to determine the values of K for a stable closedloop system. Check the answer by means of the Routh-Hurwitz criterion.
7.13.
The Nyquist
criterion
was
By
sketching the Nyquist plot of G(s)H(s) that corresponds to the Nyquist path, it is possible to tell whether the system's characteristic equation has roots in the right half of the i-plane.
(a) Define a new Nyquist path in the j-plane that may be used to ensure that
all the complex roots of the characteristic equation have damping ratios
of a closed-loop system.
7.14.
than a t
may
left
is
in Fig. P7-14.
stable.
*- as)
R(s)
Figure P7-14.
Chap. 7
Problems
(b)
7.15.
of the Re(s)
all lie
373
to the
is
shown
*,(*)
*-C.
R 2 (s)
*~C 2 (s)
in Fig. P7-15.
(s)
Figure P7-15
The
is
given by
K
(j
and
K is
IX*
2)
K so
is
asymptotically stable.
(a)
(b)
7.16.
Figure P7-1 6 shows the block diagram of a control system in which the amplifier
Amplifier
c
G(s)
Amplifier
characteristic
G(s)s(l
+0.1
,s)(l
+0.2s)
Figure P7-16.
374
Stability of Control
Systems
Chap. 7
will
be absolutely
maximum
value of
stable.
Figure P7-17 shows a control system that has a synchro control transformer as
an error transducer. The output of the synchro is proportional to the sine of the
input shaft position. Assuming that the synchro operation
n/1
limited
to
K so
is
<9 < nil, the input-output characteristic of the device may be repre-
is
Y%
J
shown
in Fig. P7-17.
absolutely stable.
Synchro
control
G(s)
transformer
G(s) =
s(\
+0.5s)(l +.s)
Synchro
10
characteristic
Figure P7-17.
8
Root Locus Techniques
8.1
Introduction
it is
a given range. Since the characteristic equation plays an important role in the
dynamic behavior of
systems theory
is
linear systems,
an important problem
in linear control
teristic
systems.
also be used to
In general, the root locus problem for one variable parameter can be defined
by referring to equations of the following form
F(s)
where
K is
s"
ai s"->
+...
a n _ lS
an
oo
and
oo.
The
coeffi-
!,...,
The root
a,
x ,
...
bm _
t ,
in this chapter.
Although the
loci
when
K varies between
-co and
375
376
Chap. 8
oo are generally referred to as the root loci in control system literature, the follow-
Root
1.
loci: the
values; that
< K<
is,
Complementary root
2.
loci: the
Root contours:
3.
A'
assumes positive
is,
oo <
K<
when
loci of roots
0.
The complete
complementary root
8.2
when
oo.
varies.
and the
loci.
is
is
the characteristic equation of a linear control system that has the closed-loop
transfer function
C(s)
~~
R(s)
The
G(s)
G(s)H(s)
left side
,
v
2>
also satisfy
1
G(s)H(s)
(8-3)
intri-
we
do not contain K, we
get
,
Comparing Eqs.
lished
K(s m
(8-3)
s-
and
+ 6, J""' + ... + b m . s + b m =
+ ,*- + ... + _,* + a.
)
(8-4)
we
(S
'
can be estab-
+ b m + bJ
w w = K(s ++ b^+...
+
+ a_i5 + a
- lS
G(s)H(s)
aiS"
s"
where G(s)H(s) is known as the loop transfer function of the control system.
Since we have mentioned that the root locus technique is not limited to control
systems, in general, given Eq. (8-1), we can regard G(s)H(s) of Eq. (8-5) as the
loop transfer function of an equivalent control system. The control system
is
Later
we
it
its
characteristic equation.
many
from Eq.
(8-1) is a
Sec. 8.2
name, we
We
now
can
Now we
oo
and oo
the conditions under which Eq.
varied between
is
(8-3) is
= KG^H^s)
G(s)H(s)
where
K as
define the complete root loci as the loci of the points in the
satisfied.
377
(8-6)
G^H^s) no longer contains the variable parameter A'. Then Eq. (8-3) is
written
G^H^s)^-^
To
(8-7)
(?,(*)#,(*)
= pL
= (2k +
/GiWH^s) = 2kn
0,
1, 2,
<
-co
/G,(j)ff,(j)
where k
<
oo
(8-8)
K>
l)w
(8-9)
K<0
(8-10)
(all integers).
In practice, the complete root loci are constructed by finding all points in
the s-plane that satisfy Eqs. (8-9)
loci are
and
(8-10),
and then
the values
of
K along
the
(8-8).
is
The
arrived at analytically.
starting point
is
(8-5)
must
first
be
written*
W W -~ K(s++
G(*\H(*\
J + z ")
+ z *)
)...(s+
Xs+p
Pl
Pn
= KG^H^s)
(s
Z 'X J
(8-11)
where the poles and zeros of G(s)H(s) are real or complex-conjugate numbers.
Using Eq. (8-11), the conditions stated in Eqs. (8-8), (8-9), and (8-10)
become
m
= rlT
\G (s)H {s)\=i
l
-co<tf<co
(8-12)
I*
and
/GMH^s) =
H +z
ls
'=i
=
We shall
first
(2k
+
is
l)7i
-^ ls+
j=i
Pj
(8-13)
<K<
a rational function of
oo
s.
378
Chap. 8
/G^HM = 2 + - S
/s
=
for
A:
/s
z,
j-
i=i
2kn
(8-14)
-oo<K<0
It
0, 1, 2,
was mentioned
and
(8-14)
may
construction of the complete root loci in the s-plane. In other words, Eq. (8-13)
implies that for any positive value of AT, a point
(e.g., Si) in
the s-plane
is
a point
on the root loci if the difference between the sums of the angles of the vectors
drawn from the zeros and the poles of G(s)H(s) to s, is an odd multiple of
180. Similarly, for negative values of A", Eq. (8-14) shows that any point on the
complementary root loci must satisfy the condition that the difference between
the sums of the angles of the vectors drawn from the zeros and the poles to
the point
To
root
is
illustrate the
loci, let
0.
us consider
G(s)H(s)
K(s + z )
+ PiKs + Pi)
t
s(s
(8-15)
The locations of the poles and the zero of G(s)H(s) are arbitrarily assumed, as
shown in Fig. 8rl. Next, we select an arbitrary point, s in the .y-plane and draw
vectors directing from the poles and the zero of G(s)H(s) to the point j,. If j,
t ,
s-plane
U+p
3 )].
[K(s
zi)]/[s(s
+ p 2)
is
379
Sec. 8.2
<K<
G(s)H(s)],
oo)
it
.\_?j_IiL
|jllkl+/2ll*l+/j|
(8-13),
ISj
(8-16)
l*|
1)tt
fc
0,
1,2,...
(8-17)
Similarly, if s t
<
0), it
must
/j
to be a point
is
satisfy
zi
= 0, 1, 2,
As shown in Fig.
(/i.
M+
loci
<K
( oo
is,
J?2
/Ji
;>, )
= Ikn
(8-18)
for&
ri ,
0,,,
d P2 and
,
0,,,
vectors measured with the positive real axis as zero reference. Equations (8-17)
and
(8-18)
become
tl
(0 Pl
P3 )
(2k
d pi )
2kn
+e +d
p!
0<K<oo
l)n
(8-19)
and
0.i
~ (K +
-co<K<0
(8-20)
respectively.
If s t is found to satisfy either Eq. (8-19) or Eq. (8-20), Eq. (8-16)
determine the value of AT at the point. Rewriting Eq. (8-16), we have
|ff
kilbi
\
If,
^x
zx
is
+ Pi\\s, + p
s + z
i
is
used to
(8
_ 21
z to
The
= B^-
(8-22)
complementary root
loci.
1.
2.
and (8-10).
The determination of the values of K at points on the root
the complementary root loci by use of Eq. (8-8).
From
search for
all
the
*,
the basic principles of the root locus plot discussed thus far,
all
it
and
may
380
and
Chap. 8
is
a very tedious task. However, aided with the properties of the root
we
(8-10)
loci that
most cases
is
with some experience on the part of the analyst, the root loci can be sketched by
following through the "rules" of construction.
the Spirule, can also be used to help plot the root locus diagram. However,
since the Spirule
is
if
we
know
already
it
can be used
loci,
effectively
when we
set
general knowledge
and
~ 29
be used
8.3
before
successfully.
The following
from the
relation between
be regarded only as an aid to the construction of the root loci and the complementary root loci, as they do not give the exact plots.
K=
Points
Theorem
The
8-1.
K=
ofG(s)H(s).
From
Proof:
Eq. (8-12),
m
GMH^s) =
I
= r^
(8-23)
J=i
As
approaches
(j =
no bearing
1, 2,
n). It is
G^H^s)
in Eq. (8-23).
or of G(s)H(s); that
is,
the sign of
K has
Sec. 8.3
Example
8-1
381
When K =
2)(j
3)
K(s
1)
(8-24)
= 0, j = 2, and s = 3. These
G(s)H (s) if we divide both sides of Eq.
Golden Rule) and establish the relation-
0,
AT (the
ship
G(s)H(s)
K(s
s(s
1)
2)C$
(8-25)
3)
Thus
K(s
G(s)H(s)
s(s
The
three
points
2)(j
1)
(8-26)
3)
loci are as
shown
in Fig. 8-2.
.s-plane
;cj
K=
K = oo
3)
K(s
1)
which
K=Q
K=
K=
loci
of s(s
2)
0.
Points
Theorem
8-2.
The
K = oo
zeros of G{s)H(s).
Proof:
Referring again to Eq. (8-23), as K approaches co, the equation
approaches zero. This corresponds to i approaching the zeros of G{s)H{s); or
">
''
^
* {i=
s approaching z
1,2, ...,m).
t
'
Example
8-2
+ 2)(s + 3) + K(s + 1) =
(8-27)
It
apparent that when K is very large, the equation can be approximated by
K(s+l) =
(8-28)
which has the root s = -1. Notice that this is also the zero of G(s)H(s) in Eq. (8-26).
Therefore, Fig. 8-3 shows the point j = 1 at which K = oo. However, G(s)H(s) in
s(s
is
infinity,
382
Chap. 8
/to
s-plane
A>
-2
other K =
points at infinity
Two
the total
3)
K(s
which
1)
K=
<*>
if
the poles
and
loci
of s(s
2)
0.
and zeros
at infinity are
K = co points are at j =
co.
Number of Branches on
A branch of the complete root loci is the locus of one root when K takes on
oo and oo. Since the number of branches of the complete root
values between
loci
must equal the number of roots of the equation, the following theorem
results
Theorem
8-3.
the greater of n
Example
8-3
loci
is
three, since
+ 2)(s + 3) + Kis + 1) =
m = 1. Or, since the equation
and
and therefore three root
roots,
is
equal to
sis
of Eq. (8-1)
is
s, it
loci.
The proof of the first statement is self-evident, since, for real coeffimust be real or in complex-conjugate pairs.
The reasoning behind the second statement is also simple, since if the poles
and zeros of G(s)H(s) are symmetrical to an axis other than the real axis in the
j-plane, we can regard this axis of symmetry as if it were the real axis of a new
complex plane obtained through a linear transformation.
Sec. 8.3
Example
8-4
383
s(s
1)(*
2)
+K=
(8-30)
Dividing both sides of the equation by the terms that do not contain
G(s)H(s)
loci
to
=
s(s
K leads
l)(.s
(8-31)
2)
shown
axis), the
in Fig. 8-4.
Notice that
complete root
loci are
.s-plane
K<0
< +-
1-1
)!
I
/
/ K<0
Axis of
symmetry
Fig. 8-4.
Root
loci of s(s
l)(s
2)
+K=
0,
of symmetry.
Example
8-5
When
is
symmetrical with
respect to a point in the j-plane, the complete root loci will also be
symmetrical to that point. This is illustrated by the root locus plot of
s(s
as
shown
in Fig. 8-5.
2)(.s
+ +/X* +
1
~j)
+K=
(8-32)
384
Chap. 8
s-plane
Fig. 8-5.
Root
loci of s(s
2)(s
+y")0
-y)
A: == 0,
showing
The
co)
properties of the complete root loci near infinity in the j-plane are
important, since
m\ of the
loci will
approach
infinity in
the j-plane.
Theorem 8-5. For large values ofs, the root loci for
straight lines or asymptotes with angles given by
ek
= (?KDz
n m
(8-33)
K<
2kn
9,
where k
0, 1, 2,
According
,\n
m\
values to k.
m\ asymptotes
(8-34)
1.
0,
1, 2,
we need
However,
to assign only
Sec. 8.3
385
Equations (8-33) and (8-34) imply that the asymptotes of the complementary
root loci are linear extensions of the asymptotes of the root loci, and vice versa.
When the asymptotes of one type of root loci are determined, those of the other
type are found without further calculation.
Proof:
s^
We
b.s"-
...
bm .
bm
then have
... + a-iJ + a.
+ !?"
=
+ 6,5' + ... + b m ^s + b n + K
S"
j"
(8-35)
Carrying out the fraction of the left side of Eq. (8-35) by the process of
long division, and for large s neglecting all but the first two terms, we have
m
s"-
(a l
- b )s n m
l
^ -K
(8-36)
= (-*)"<-
(8-37)
or
The
factor
sion,
[1
and Eq.
(a,
-b
if
only the
'
'
)/sy /l
"- m>
+ (
let s
in Eq. (8-37)
is
two terms
first
+ jco,
t"vl/(n-m)
= (-*)
m)s
(8-38)
Now
becomes
(8-37)
Again,
we
get
'
is
written
o +jco
for
<K<
ai
oo,
bi
m
cos
-oo
<
(2k
n
sin
+ \)n
m
(8-40)
1/( "~ m)
I
2kn
cos
= 0, 1, 2,
and imaginary parts of both
sm
2kn
-^
n m
(8-41)
A:<0, and&
Equating the
for
+ l)n
m +
and
I
for
(2k
0< K<
real
we
have,
oo,
^i
~ Ki/<-m) cos-(2k +
,
\)n
(8-42)
and
<a^*'
Solving for AT I/( "- m) from Eq. (8-43),
/(
-->sin
+1) *
(8-43)
we have
b
m
2k+\
n cos 2k+ n
sin
n m
n
m
o
(2 *
a>
a
n
(8-44)
386
Chap. 8
or
CO
~ tan
2k
+ V* + ZUZi)
m (g
n m I
\
(8-45)
is
of
form
co
where
= M{a - a
From
(8-46)
and a
\n
is
we have
M = tan?^iirc
n m
0, 1, 2,
axis.
m\
1,
(8-47)
and
a,
M
b,
(8-48)
m
Note that these properties of the asymptotes are
only.
from Eq.
Similarly,
loci(-oo
(8-41)
we can show
< K<G),
M = tan -^n m
k
0, 1, 2,
obtained for
\n
m\
1,
(8-49)
as in
Eq. (8-48)
is
Eqs. (8-33) and (8-34), have been proved. This proof also provided a byproduct, which is the intersect of the asymptotes with the real axis of the
j-plane,
and therefore
m\ asymptotes of the
Theorem 8-6. (a) The intersection of the 2\n
plete root loci lies on the real axis of the s-plane.
(b) The intersection of the asymptotes is given by
~
where a u b u
Proof:
n,
and
fc
'
-g
m
'
com-
(8-50)
The proof of
it
that the complete root loci are symmetrical to the real axis.
The proof of (b) is a consequence of Eq. (8-48). Furthermore,
a function G(s)H(s) as in Eq. (8-5), Eq. (8-50) may be written as
if
we
define
xZL aA
b_
(8
' 51
Sec. 8.3
since
=
=
b
=
a,
sum of the
roots of s"
sum of the
finite
a,s"' x
a_,s
a.
=
(8-52)
sum
poles of G(s)H(s)
of the roots of s m
b,s m
b^^s
bm
=
(8-53)
sum of the
zeros of G(s)H(s)
finite
Since the poles and zeros are either real or complex-conjugate pairs, the imagi-
nary parts always cancel each other. Thus in Eq. (8-51) the terms in the summations
may
respectively.
It
complementary root
Example
8-6
is
2)
loci.
4)(s 2
2s
K(s
1)
(8-54)
<*'>*
The pole-zero
configuration of G(s)H(s)
is
w VL
shown
<8
+ 2)
From
in Fig. 8-6.
55 >
on the construction of the complete root loci described so far, the following information
concerning the root loci and the complementary root loci of Eq. (8-54) is obtained
1
K=
The
K=
points
of
2.
3.
G(s)H(s): s
0, s
<x>.
4.
5.
>
The angles of
byEq. (8-34):
d,=~ =
180
92
900
=^=
20O
60
A:
=0
=^ = o
6^^ =
=^ =
2
120
24O
388
Chap. 8
s-plane
Asymptotes of
complementary root
\
loci
\
Asymptotes of the complete root
Fig. 8-6.
K(s
1)
loci of s(s
4)0 2
-r
Is
2)
0.
The asymptotes of the complementary root loci may be obtained by extending the asymptotes of the root
6.
The
<r,
six
= S finite poles
- (o-4-i
of G(s)H Q)
8-7
loci intersect at
S finite zeros
+yi -l -yi)-(-i)
4-1
Example
loci.
of G(s)H(s)
_ _a
(8-56)
as
shown
in Fig. 8-6.
different equations
are
{K >
8-7. (d)
Root
may
loci:
On a
be found in the section only if the total number of real poles and
zeros of G(s)H(s) to the right of the section is odd.
(b) Complementary root loci: On a given section of the real axis, complementary root loci (K
0) may be found in the section only if the total number of
real poles and zeros of G(s)H(s) to the right of the section is even. Alternatively,
0)
<
Asymptotes of
/CO
i-plane
s-plane
Asymptote of
root locus
Asymptote of
complementary root locus
"l
'l
="Pi/2
/|V
_~(Pi + P 2 +P3) ,
4
45
/
/
\
I
G(s)H(s) =
s(s
s-plane
+p
G(s)H(s) =
x
Asymptotes of
complementary
\
'
/CO
s-plane
'
root loci
i\
o,
'
\|
\
/
\ vX/\
root loci
+P2+P3)^*7V^
45
r
45
\
\
/
/
K
s
G(s)H(s)=
/\
Asymptotes of Asymptotes of
complementary!
root loci
^ Asymptotes
of
'(Pi
/"
root loci
-v ->
K(s + zi)
G(s)H(s)
Fig. 8-7.
i 2(i
+p
)( J
+ p 2 )( s + p 3
loci.
389
390
Chap. 8
we can state that complementary root loci will be found in sections on the real axis
not occupied by the root loci.
In all cases the complex poles and zeros ofG{s)H{s) do not affect the existence
properties of the root loci on the real axis.
Proof:
1
The proof of
the theorem
is
At any point (e.g., s ) on the real axis, the angles of the vectors drawn
from the complex-conjugate poles and zeros of G(s)H(s) add up to
t
Only the
point
3.
St
real poles
and zeros that lie to the left of the point contribute zero degrees.
Each real pole of G(s)H(s) to the right of the point j, contributes
180 and each zero to the right of the point contributes 180 to
Eqs. (8-13) and (8-14).
The last observation shows that for s to be a point on the root loci, there
must be an odd number of poles and zeros of G(s)H(s) to the right of s u and for
Si to be a point on the complementary root loci the total number of poles and
t
s-plane
Complementary
root loci
Root
loci
on the
real axis.
Sec. 8.3
391
Example
8-8
complete
two
loci
root loci
on the
complementary root
shown in
No-
loci.
Angles of Departure (from Poles) and the Angles of Arrival (at Zeros)
of the Complete Root Loci
{arrival)
ofG(s)H(s) denotes the behavior of the root loci near that pole (zero). For the root
loci
(K >
0) these angles
(8-13).
-<*><-
it is
For
instance, in
desired to deter-
AT
Fig. 8-9.
Complete root
loci of s(s
3)0 2
2s
2)
+ K=
to illus-
392
Chap. 8
which the root locus leaves the pole at 1 +jl. Notice that
is measured with respect to the real axis. Let us assume
that s is a point on the root locus leaving the pole at 1 + j 1 and is very near
the pole. Then s must satisfy Eq. (8-13). Thus
the
unknown
at
angle 9 2
/GjsjHjSj)
Since s,
is
-(0,
+ +
+ j\,
62
We can
92
90
26.6)
0)
(2k
1)180
(8-57)
drawn
becomes
(2k
(8-58)
is
obtained for
all
92
which
is
the
same
-431.6
as 71.6.
now
of G(s)H(s)
same point
is
determined,
differs
from
this
be used.
The points where the complete root loci intersect the imaginary axis of the
and the corresponding values of K, may be determined by means of the
Routh-Hurwitz criterion.
For complex situations with multiple intersections, the critical values of
K and co may be more easily determined approximately from the Bode diagram.
s-plane,
Example
8-9
drawn
loci
3)(s 2
of the equation
+2s +
2)
+K=
(8-59)
in Fig. 8-9.
may
occur.
A root locus diagram can, of course, have more than one breakaway point.
Moreover, the breakaway points need not always be on the real axis. However,
because of the conjugate symmetry of the root loci, the breakaway points must
either be real or in complex-conjugate pairs.
Sec. 8.3
393
s-plane
s-plane
K=Q
<
Breakaway
Breakaway
point
point
(b)
(a)
Fig. 8-10.
/'co
s-plane
breakaway point.
Because of the symmetry of the root loci, it is easy to see that the root loci
and (b) break away at 180 apart, whereas in Fig. 8-11 the four
< oo)
root loci depart with angles 90 apart. In general, if n root loci ( oo <
in Fig. 8-10(a)
394
Chap. 8
Several graphical and analytical methods are available for the determination
Two
Method
KG
1.
loci
of 1
dG^sJH^s)
Proof:
(8-5)
may
(86Q)
(5)
is
KP(s)
be written
GW" =
we
Then Eq.
If
analytical
varied by an increment
Q(s)
(K +AK)P(s)
(8-61)
*
(8 62)
AK, Eq.
(8-61)
becomes
KP(s),
(8-63)
we have
AKF(s)
(8-65)
where
F(S)
Since the denominator of F(s)
is
the
S)
same
a breakaway point of n
where
loci,
is
which corresponds to
=
t
(*-<>V
Q(s)+ KP(s)
(F*i? = (0r
<
8 - 67 >
a constant.
1+^ =
(8-68)
Taking the
have
limit
on both
sides
of the
lim
Ajc-o
last
* =
As
equation as
AK approaches
zero,
^=
we
(8-70)
v
/
as
We have shown that at a breakaway point on the root loci, dKjds is zero.*
*The quantity (dsjs)l(dKIK)
is
away
it is
proved that
at the break-
Sec. 8.3
Now,
must
395
also satisfy
+ KG^H^s) =
(8-71)
or
it is
is
equivalent to
ds
It is important to point out that the condition for the breakaway point given
by Eq. (8-73) is necessary but not sufficient. In other words, all breakaway points
must satisfy Eq. (8-73), but not all solutions of Eq. (8-73) are breakaway points.
To be a breakaway point, the solution of Eq. (8-73) must also satisfy Eq. (8-71);
or,
All real solutions of Eq. (8-73) are breakaway points on the root
loci
( oo
<K<
The complex-conjugate
points only
if
of the s-plane
is
loci.
solutions
of Eq.
(8-73)
are
breakaway
cause difficulty in the effective use of Eq. (8-73), since the other
properties of the root loci are usually sufficient to provide information
trial
and
error.
Example 8-10
Consider that
it is
loci.
equation
s(s
2)
K(s
4)
(8-74)
Based on some of the theorems on root loci, the root loci of Eq. (8-74) are easily
shown in Fig. 8-12. It can be proven that the complex part of the loci is
described by a circle. The two breakaway points are all on the real axis, one between
and 2 and the other between 4 and -co.
sketched, as
of Eq. (8-74) by
s(s
2) (the
the identity
C,(5)ff,(5)
= g^
(8-75)
dG&Wiis)
= *fr +
2)
or
s2
we
Ss
+ l)(s + 4) =
+ 2)
=
2(s
s 2 (s
ds
(8-77)
loci are at s
396
Chap. 8
s-plane
Root
Fig. 8-12.
loci
Note also
of s(s
2)
K(s
4)
0.
that the
all
on
Example
8-11
2s
+ K(s +
2)
(8-78)
G(s)H(s)
s2
(8-78)
by s
2s
2,
2)
+2s +
(8-79)
Based on the poles and zeros of G(s)H(s), the complete root loci of Eq. (8-78)
shown in Fig. 8-13. The diagram shows that both the root loci and the
complementary root loci possess a breakaway point. These breakaway points are
determined from
are sketched as
<iWgi()
ds
...
d
ds s 2
s
(s
2s
2)
(8-80)
2)
Sec. 8.3
397
/CO
x-plane
Fig. 8-13.
Root
loci of s 2
2s
K(s
2)
0.
or
s2
Upon
and
loci,
4s
(8-81)
whereas s
Example
8-12
0.586
is
3.414
is
0.586
loci.
+ 4)(s 2 + 4s +
20)
+K=
(8-82)
Dividing both sides of Eq. (8-82) by the terms that do not contain K,
1
G(s)H(s)
K
s(s
4)(s 2
+4s +
we have
(8-83)
20)
ddJsW^s)
ds
to
s,
we
get
+ 24s 2 + 12s + 80 _
+ 4)(s +4s + 20)] ~ u
4s 3
[s(s
(8 " 84)
or
s3
6s 2
18s
20
(8-85)
398
Chap. 8
x-plane
Fig. 8-14.
Complete root
loci of s(s
4)0 2
4s
20)
+K
=--
0.
Because of the symmetry of the poles of G(s)H(s), one of the breakaway points is
determined to be at s = 2. The other two breakaway points are found by
solving Eq. (8-85) using this information; they are s = 2 +j2.45 and s = 2
easily
j2A5.
Example
8-13
In this example
we
shall
necessarily represent
are
shown
show
loci of the
2s
equation
+ 2) + K =
(8-86)
breakaway point
in this case.
+ KGMH^s) =
do not
loci.
+ s(s 2 +
K
2s
+ 2)
loci
have any
(8-87)
Sec. 8.3
Fig. 8-15.
Complete root
(8-60),
loci
of s(s*
2s
+K=
2)
399
0.
we have
1
ds s{s 2
ds
+25 +
(8-88)
2)
which gives
3s 2
4s
(8-89)
The roots of Eq. (8-89) are s = -0.677 +./0.471 and* = -0.677 -yO.471. These two
roots do not represent breakaway points on the root loci, since they do not satisfy Eq.
(8-86) for any real values of K. Another way of stating this is that Eq. (8-89) is not a
factor of Eq. (8-86) for any real K.
Method
( co
2.
An
<K<
oo)
400
Chap. 8
Let the equation for which the root loci are desired be written as
F(s)
=A
s"
A,s-
Arrange the
Bo
Al
Si
B. 2 s
coefficients of F(s)
Ap
Form
4.
Am
-+
(8-90)
i ,
= Bvs"- + B,5"- +
+
where B = nA B = (n 1)A U
F'(s)
3.
A n _ lS
2.
and
...
An- 1
B2
...
U_i
B_,
(8-91)
etc.
F'(s) in
A2
An
equation, that
is,
for n
is
4.
extended.
s*
Ap
Ai
s*
Bi
BoAi
BiAo
2
s
S2
s2
,i
,i
n
Do
E =
BqC\
BiCo
To
DqBi
DiBp
Dp
<>
Fo
sp
_ BpC B
= D B3 - D 3B
= D B2 - D 2 Bq
Dp
Fi
Am
Bi
n
Ul
Cp
To
Dp
n3
B
_
~~
Ei
Bo
OCp
ft
ft
=0
D2
Di
F = D Ei ~ DjE
G = F Di D Fl
^oG
ff0==
To
'
Dp
j>
BqCi
B 2 Cp
Ho
B2
Bi
r _
At
B,
Bq A*
Bo
Bo
Bo
r _ B A 3 B 3 Ap
n _ BpAi BjAg
Bo
j3
A2
B2
= DoEi ~ DlE
F2 =0
F* D
Gi= F D *'o
Ft
-F Go
1
1.
The
row of the
tabula-
Sec. 8.3
2.
3.
The
1.
If
we
same
(n
401
1),
row of a group
same group.
it is
first
row of the
If F(s) has multiple-order roots, which means that the root loci will have
breakaway points, a row of the tabulation shown above will contain all zero
The multiple-order
roots,
row of zeros.
Let us use the following numerical examples to illustrate the application
of the tabulation method of finding the breakaway points.
Example 8-14
F(s)
(s
1) C$
2)
s*
6i 3
13s 2
12s
(8-92)
We have stated the problem in such a way that the equation is known to have two
double roots at s = 1 and s = 2. We are merely going to demonstrate the propof the tabulation when the multiple-order-root situation occurs.
erties
= 4s +
3
F'( s )
l&s
26s
s3
4
(4)(6)
S3
4
4
S2
_i
S2
6
_i
S2
is
(4X13)
made
18
_3
in the following
12
26
13
(D(26) _
(8-93)
13
18
(D(18)
we have
12
s,
(4X12)
4
26
manner:
12
0X12)
4
12
-i
18
12
_3
"i
jl
is
completed,
is
-\s 2
- Is -
(8-94)
or
s2
The
multiple order.
= 1
+2=
s = 2,
3s
and
(8-95)
402
Example
Chap. 8
8-15
Let us consider a root locus problem, that is, the determination of the
breakaway points on the root loci. The equation, Eq. (8-74), con-
Example 8-10
sidered in
F(s)
=j +
2
(2
will
K)s
+ 4K =
is
rewritten
(8-96)
Then
F'C?)
The following
tabulation
2s
(2
+ K) =
J2
sl
+K
2
-< 2 +4
+K
+K
4K
4K
2
2
si
(8-97)
made
is
+K
*> 2
to contain
all
we
set
4K- (2
4-
JQ 2
(8-98)
4=
(8-99)
or
-K +
2
Solving for
12^:
and
K= 11.656
When
contain
all
two
values, the s
row of
row preceding
the s row.
The equation
thus formed
+ (2 + K) =
K = 11.656 into Eq.
2s
Now substituting K =
is
0.344 and
is
(8-100)
(8-100) in turn,
we find
the
loci at
-1.172
K = 0.344
-6.828
K= 11.656
and
apparent that these answers agree with those obtained in Example 8-10. Furthermore, a by-product of the tabulation method is that the values of K at the breakaway
points are also determined. In fact, the values of K are always determined first before
the breakaway points are found.
It is
Example 8-16
Now
(8-82)
written as
= s* +
K=
(8-101)
The root loci of Eq. (8-101) have three breakaway points at s = 2, 2 + J2.45, and
2 y'2.45. Let us now determine the breakaway points by the tabulation method.
F(s)
Ss 3
+36s 2 +80s +
Sec. 8.3
We
403
have
= 4s + 24s +
12s
+ 80 =
(8-102)
18s
(8-103)
is
F'( s )
6s 1
20
made
36
80
18
20
18
60
18
20
24
6
2
isT
- K-40
18
-40
20
#-40
24
10- #-40
#-40
20
We would like to interrupt the tabulation at this point to remark that the
elements
row of the
s1
equation
6s 2
24s
- 40) =
(K
K=
100
(8-104)
or
s
4s
10
(8-105)
= 2 +y'2.45 and
which are the solutions of Eq. (8-105).
In order to complete the tabulation, we must now consider that
Therefore, the two breakaway points occur at s
row
(the second
row of
= 2 y'2.45,
K^
100, so that
finite. In
100 are already determined, should there
be any additional multiple-order roots, they would have to occur at different values of
K than 100. Therefore, in the remaining three rows of tabulation it is implied that
fact, since
K ^
100.
Resuming
the tabulation,
K=
we have
#-40
#-40
20
12
10
_ #-40
6
KNow
the only
row
that can be
is
64
all
zero
is
the s row,
found by substituting
10
K^ 100
K=
^)-+(
K=
64. Therefore,
64 into
40
')=
(8-106)
which gives
s
An
factor 10
alternative
(#
= -2
the tabulation
is
common
as follows
404
Chap. 8
K-40
12
K-64
Therefore, the
Example
8-17
same
results,
K=
64 and s
= 2,
we
show
In this example
shall
are obtained.
Equation (8-86)
is
method actually
Example
in Fig. 8-15,
points.
written
F(s)
Then
F'(s)
The following
diagram
tabulation
s3
+ 2s +
3s 2
4s
2s
+K=
(8-107)
(8-108)
made
is
K
2
K
2
K-%
1
- 21K
K-
*-(*--)('-)
It is
we
last
set
2-\(k-)(i-^)=0
which
is
(8-109)
simplified to
811s: 2
real values of
- 4SK + 16 =
K that will satisfy this
(8-110)
is that F{s) does not have any multiple-order roots, or the root
have any breakaway points.
conclusion
loci
do not
The condition
sary but not
sufficient.
The method
breakaway point
is
neces-
may be
method
Sec. 8.3
405
These methods also represent ways of solving for the multiple-order root
of an equation.
Calculation of
K on
Once the root loci have been constructed, the values of K at any point s^
on the loci can be determined by use of the denning equation of Eq. (8-8); that
is,
|r|
If G,(j)#i(j) is of the
1*1
form shown
iwwi
in Eq. (8-1
1),
<!M1,)
Eq. (8-12)
is
written
ni*i+/j
^
(8-112)
or
l-i
(8-113)
Usually,
method
is
+ 2s + 2 + K(s + 2) =
The value of K at the point s,
s2
are
shown
in Fig. 8-16.
(8-114)
is
given by
*=^
(8-115)
where A and B are the lengths of the vectors drawn from the poles of G(s)H(s)
= K{s + 2)/(s 2 + 2s + 2) to the point j, and C is the length of the vector
drawn from the zero of G{s)H(s) to s In the illustrated case s is on the root
t
loci,
so
K is positive.
have a negative
If j, is
loci,
K should
sign.
The value of at the point where the root loci intersect the imaginary axis
can also be found by the method just described. However, the Routh-Hurwitz
criterion usually represents a
more
direct
method of computing
this critical
value of K.
+ Pl )(s +p
)...(s+
p)
K(s
z t )(s
z2 )
(s
zm )
(8-116)
Then, except for extremely complex cases, these rules are usually adequate for
the analyst to make a reasonably accurate sketch of the root loci just short of
plotting them point by point. In complicated situations, one has to rely on
a computer as a more practical means of constructing the root loci.
406
Chap. 8
.s-plane
method of finding
the values of
serves as an illustration
Example
8-18
+ 6)(s 2 + 2s + 2) + K(s + 3) =
( oo < K < co) of the system are to be
loci
5)(s
(8-117)
constructed. Using
the rules of construction, the following properties of the root loci are determined:
1.
j1,
G(s)H(s)
K(s
=
s(s
2.
5)(s
3)
6)(s 2
+2s +
(8-118)
2)
K=
oo, oo,
3.
4.
real axis of
the s-plane.
5.
The
angles of the asymptotes of the root loci at infinity are given by [Eq.
(8-33)]
6k
for
= 0,
j-plane as
= SJ3
= Q!LTE
n m
5 -1
1, 2, 3.
K approaches
o<*<co
(8-119)
approach
infinity in the
loci that
angles
Sec. 8.3
407
of 45, -45, 135, and -135, respectively. The angles of the asymptotes
of the complementary root loci at infinity are given by [Eq. (8-34)]
ek
Therefore, as
approach
6.
The
2kn
2kn
-co
1
<K<0
(8-120)
K approaches
infinity
is
The
results
from these
(8-121)
;w
s-plane
-n
K=
+-K
K=
AT= +
-6
oo\ y'45A
-3' \~
K=
_1
\\ K
-2.5 \
=
X
--/l
2s
2)
K(s
3)
0.
5)(s
6)(s 2
AT->--
408
Chap. 8
more than the behavior of the root loci as s < oo. The
complete root loci portions can be correctly sketched as shown in Fig.
8-17 only if the entire solution to the root loci problem is known.
Complete root loci on the real axis There are root loci (0
K < oo)
cate nothing
7.
<
between s =
There are complementary root
loci
between
on the
and
and
real axis
-co
is,
( oo
s
3,
5, and s = 6.
0) on the remaining
and s = 5, and * =
s ==
< K<
= 3
K=
s-plane
Complementary
root loci
Complete root
K(s + 3) = 0.
Fig. 8-18.
2)
8.
loci
on the
5)0
6)0 2
2s
the pole at
1 +jl, and
St is
very close to
-1
+jl
as
shown
in
+3 -
(/j,
Is,
+yl +
/s t
1
+5 +
-;1)
sv
/.
(Ik
1)180
i-plane
Fig. 8-19.
(s
6)0 2
loci
of s(s
5)
(8-122)
Sec. 8.3
409
or
26.6
for
A:
= 0,
(135
90
14
+ 9) s;
11.4
(2k
1)180
(8-123)
1, 2,
Therefore,
9 =* -43.8
(8-124)
Similarly, Eq. (8-14) is used to determine the angle of arrival of the complementary root locus arriving at the point 1 +jl. If this angle is
designated as
9',
it is
9'
9.
The
180
43.8
136.2
13s*
54s 3
82s 2
(60
is
determined by
rewritten
+ K)s + 3K =
(8-126)
60
47.7
+K
60
54
82
is
is
13
65.6
is,
(8-125)
3K
0.769-ST
3K
0.212A:
3940- 105A:-0.163*:2
65.6
0.212K
3K
.s
For Eq.
(8-126) to have
same
sign.
- 0.212#>
- 105K - 0.1 63^ >
or
65.6
3940
or
K<
K<
K>0
Hence
309
(8-127)
35
(8-128)
(8-129)
all
the roots
K lies between
A(s)
Substituting
K=
(65.6
- 0.212A> + 3K =
2
which
105
we have
(8-131)
yields
s
10.
(8-130)
/1.34
trial
410
Chap. 8
it is
since
we know
Applying dK/ds
s3
error,
13.55*
66.s
142s 1
123s
45
After a few trial-and-error calculations, the root of the last equation that
From
is
is
found to be
5.53.
the information obtained in these 10 steps, the complete root locus diagram
Fig.
8-20.
3)
0.
Complete root
loci
of s(s
5)(s
6)0 2
2s
2)
K(s
Sec. 8.3
Of course,
minor prop-
of the root loci which are not mentioned here. However, in general,
Table 8-1
K=
Table
K=
The
points
them.
points
the poles of
at
at
infinity.)
2.
K=
oo
points
The
K = oo
points
loci are
infinity.)
3.
Number
of separate
root loci
4.
The
Symmetry of root
loci
transfer functions
loci
loci is
s-
plane.
5.
Asymptotes of root
loci as i
>
oo
For
large values of
for the
complementary root
a
Ok
where
Intersection of the
s,
(a)
asymptotes
A:
The
0,\,2,...,\n
(b)
a
on
loci
(K
<
0)
2kn
The point of
m\
\.
is
given by (for
2 real parts of
loci
0) are
Root
>
on
(centroids)
7.
(K
and
6.
the
real axis
all
values of
_ 2 real parts
poles of G(s)H(s)
on
K)
of
zeros of G(s)H(s)
On
K>
total
(^< 0)
8.
Angles of departure
and
arrival
it is
1 1
>
and which
is
8-1.
412
Chap. 8
and applying
the equation
m
/GfriV/fri)
=
The
-
1=1
Ai
(2k
z,
\)n
- 2
=
+ P;
Ai
1, 2,
0,
determined from
is
Intersection of the
= S Ai + -
+/>,;=i
j=i
= 2A:ji
= 0, 1, 2,
The values of co and K at the crossing points
imaginary axis
/G(ii)H(si)
z,-
,/s
fc
9.
The Bode
10.
...
plot of G(s)H(s)
may
of the
also be used.
Breakaway points
The breakaway
(saddle points)
points
may
criterion.
loci
0,
These are necessary conditions only. Alternatively, the breakaway points are
determined from a tabulation using the coefficients
of the characteristic equations F(s) --=
and F'(s)
= 0. The conditions are necessary and sufficient.
or dG(s)H(s)/ds
1 1
The
Calculation of the
values of
K on the
absolute value of
root loci
0.
K at any point s
on the corn-
is
1
\G(si)H{ Sl )\
product of lengths of vectors drawn
s\
8.4
Although the root locus diagram is primarily intended for the portrayal of
the trajectory of roots of a polynomial when a parameter, K, varies, the technique
may
it is
The
is
best illustrated
when
all
by an example. Con-
F(s)
To
principle
s3
3s 2
45
20
(8-132)
form of Eq.
first
convert Eq.
(8-4).
we can regard any one of the four coefficients of the polynomial as the parameter
number of these coefficients as K will
see that
it is
Thus Eq.
more
of the root
loci.
In this case
it is
not
difficult to
(8-132) leads to
3s 2
Sec. 8.4
+
which
is
A* + 3)
with K = 4.
413
(8-133)
Furthermore,
it is
apparent that
must also satisfy Eq. (8-133). Now the problem of solving Eq. (8-132) becomes that of a root locus problem, based on the pole-zero
the roots of Eq. (8-132)
configuration of
bWW _- Ks\s ++
G(s)H(s)
(s
5)
(8-134)
3)
The desired roots are then found by setting K = 4. From a logistic standpoint,
we have embedded a specific problem in a more general one by first solving for
the solution to the general problem.
The root locus diagram based on the function G(s)H(s) of Eq. (8-134) for
K is shown in Fig. 8-21. When K = 4, the real root of Eq. (8-132) lies
between 3 and 5, while the other two roots are complex with positive real
positive
s-plane
K-5
=
K=
Fig. 8-21.
4s
-*
K=
20
K=
0.
3s 2
414
Chap. 8
Example
s*
its
is
0.2475
+ y2.381
and
8-19
Since this
are found to be at s
5s 3
2s 2
a fourth-order polynomial,
roots will be
more
difficult
by the
+s+
it is
10
(8-135)
first
divide both
first
2(s 2
+ 0.5j +
+ 5)
5)
s*(s
"
(8_136)
or
G(s)H(s)
^afr+y
5>
K=2
(8-137)
The root locus diagram based on the poles and zeros of G(s)H(s) is constructed as
0. However, from this root locus diagram it is not
shown in Fig. 8-22 for oo > K
>
s-plane
Fig. 8-22.
2s 2
+s+
10
= 0.
5s 3
Sec. 8.4
clear
the
first
The root
(8-135) by
s 2 (s 2
+ 10
+ 5s +
(8-138)
2)
locus plot of
G(s)H(s)
with co
415
we have
s
> K>
shown
+ 10)
+ 5s +
K(s
s 2 (s 2
2)
K=
(8-139)
is
is
Eq. (8-135).
One
(8-135) by (s
real root
1.575),
is
found to be
we have
s
Now we
in Fig. 8-23.
at s
the remainder,
3.425.S 2
3.394.T
6.346
(8-140)
repeat the procedure to find the roots of Eq. (8-140). Since one of the
we can reason
that
it
must be to the
right of s
/CJ
s-plane
Fig. 8-23.
2s 2
Root
+s+
10
0.
5s 3
4.56 on
416
Chap. 8
this
G(s)H(s)
-co
-3.3940?
s 2 (s
K(s
s 2 (s
1.87)
3.425)
- 1.87)
+ 3.425)
-3.394
(8-141)
loci
< K<
'
/w
s-plane
K<0
K<0
Fig. 8-24.
3.3945
6.346
3.425.S 2
0.
diagram and the root loci of Fig. 8-24 reveals that the real root must lie between 3.425
and 4.56. To find the real root of Eq. (8-140), which is at s = 4.493, is a simple
matter. The two complex roots are subsequently found to be at 5 = 0.534 + j 1.057
and s = 0.534 / 1.057. Thus the roots of Eq. (8-135) are now all determined:
5
s
s
.y
=
=
=
=
-1.575
-4.493
0.534
+/ 1.057
0.534
-y 1.057
(8-142)
Some
Sec. 8.5
where F(s)
is
a polynomial in
form
s.
is
417
(8-142) as
+ GO) -
P(s)
(8-143)
Then, G(s)H(s) = Q(s)/P(s), or G(s)H(s) = P(s)/Q(s). The selection of the polynomials P(s) and Q{s) from F(s) is more or less arbitrary, although in general
the orders of P(s) and Q{s) should be close so that the root locus problem is made
how
this is
done.
In essence, the
method of root
is
that of utilizing
some
of the roots. The basic method of finding the exact roots is still cut-and-try.
For high-order equations, and for equations only with complex roots, the root
locus
still
8.5
method of root
finding
may
is
aspects of the root locus techniques is that for most consystems with moderate complexity, the analyst or designer may conduct
a quick study of the system in the s-plane by making a sketch of the root loci
using some or all of the rules of construction. In general, it is not necessary to
trol
make an exact plot of the root loci. Therefore, time may be saved by skipping
some of the rules, and the sketching of the root locus diagram becomes an art
that depends to
In this section
loci
in the
and Zeros
In Chapter 6 the effects of the derivative and integral control were illustrated
From
when
we may
may
state that
tion G(s)H(s) in the left half of the s-plane has the effect of pushing the original
it is
we can
difficult to
make
a precise
by several
examples.
Let us consider the function
G(s)H(s)
The zeros of
8-25(a).
K
,
>
(8-144)
These root
OO
t
s-plane
s-plane
K=
K=Q
<
K
i
(b)
(a)
(d)
(c)
Fig. 8-25.
G(s)H(s).
418
that
show
Some
Sec. 8.5
and
= a. Now
let
us introduce a pole at
G(s)H(s)
s(s
a)(s
~|-
419
-b so that
b>
b)
(8- 1 45)
Figure 8-25(b) shows that the additional pole causes the complex part of the
The angles of the asympchanged from 90 to 60. The breakaway point is also moved to
the right. For instance, if a =
and b = 2, the breakaway point is moved from
0.5 to 0.422 on the real axis. If G(s)H(s) represents the loop transfer funcroot loci to bend toward the right half of the s-plane.
totes are
tion of a feedback control system, the system with the root loci in Fig. 8-25(b)
may become
unstable
if
the value of
K exceeds
The system
is
now
loci
The angles of
two
stability condition
loci are
moved
loci are
45.
even more restricted. Figure 8-25(d) illustrates that the addition of a pair of
complex-conjugate poles to the original two-pole configuration will result in
Addition of zeros. Adding zeros to the function G(s)H(s) has the effect of
the root loci toward the left half of the j-plane. For instance, Fig. 8-26(a)
moving
OO
/CJ
/CO
s-plane
6 =
is
<
5-plane
K =
--^
/
/
\
\
\
\
-o
a:
= o
tf
-*
*
)
K=0
K=
-a
~b
-a/2
-a/2
1
K=
/
'
'
K =
b = <*>^
%
i
(a)
Fig. 8-26.
G(.s)H(s).
(b)
that
show
420
Chap. 8
71
K/7
f
JCO
<
s-plane
4
1
u
\
AT
K=Q K =
A'
\
\
\\
w
A
-\
(c)
that
show
azerotoG(.s)//(.f).
and form a
circle.
>
Therefore,
if
at 5
= b
is
is improved
by the addition of the zero. Figure 8-26(b) illustrates that a similar effect will
result if a pair of complex-conjugate zeros is added to the function of Eq. (8-144).
Figure 8-26(c) shows the root locus diagram when a zero at v = c is added to
the transfer function of Eq. (8-145).
Effects of
It was mentioned earlier that the construction of the root locus diagram
depends greatly on the understanding of the principle of the technique rather
than just the rigid rules of construction. In this section we show that in all
cases the study of the effects of the movement of the poles and zeros of G{s)H(s)
on the root loci is an important and useful subject. Again, the best way to
illustrate the subject is to
II
u,
e
c
:=;
c
"E.
o
II
.a
S
^
"
II
">
II
"
O ,
-o
S3
f}
o ,
aS O Q
II
as a> 00
*,
.fa tt
II
421
o
II
422
Some
Sec. 8.5
423
/CJ
5-plane
K=
K=
(e)a=
Example 8-20
^(s
which
is
easily
a)
G(s)H(s)
K(s
1
--=
b)
(8-146)
G(s)H(s)
= 0,
with
*^
(8-147)
= 1 and investigate the root loci of Eq. (8-146) for several values of a.
Figure 8-27(a) illustrates the root loci of Eq. (8-146) with a = 10 and b
1. The
two breakaway points are found at s = 2.5 and 6.5. It can be shown that for arbitrary a the nonzero breakaway points are given by
Let us set b
-1-
^Vo 2
10a
(8-148)
= 9, Eq. (8-148) indicates that the breakaway points converge to one point
3, and the root locus diagram becomes that of Fig. 8-27(b). It is interesting
to note that a change of the pole from 10 to 9 equals a considerable change to the
root loci. For values of a less than 9, the values of s as given by Eq. (8-148) no longer
satisfy the equation in Eq. (8-146), which means that there are no finite, nonzero, breakaway points. Figure 8-27(c) illustrates this case with a = 8. As the pole at s
a is
When
at j
------
424
Chap. 8
moved
farther to the right, the complex portion of the root loci is pushed farther
toward the right-half plane. When a = b, the pole ats = a and the zero at b cancel
each other, and the root loci degenerate into a second-order one and lie on the imaginary axis. These two cases are shown in Fig. 8-27(d) and (e), respectively.
Example
8-21
s(s 2
which
is
+
+
2s
G(s)H(s)
a)
K(s
G(s)H(s)
+ 2) =
= 0, with
+
(8-149)
(8-150)
l a)
The objective is to study the complete root loci (-co < K < co) for various values of
a{> 0). As a start, let a = so that the poles of G(s)H(s) are at s = 0, 1, and 1.
s{
this
By
setting dG(s)H(s),lds
leads to
3
.y
4s 2
As
(8-151)
loci
with a
1.12; that
is,
Since the real parts of the poles and zeros of G(s)H(s) are not affected by the value
of
a,
By
1.12 are at s
is
always at the origin of the .j-plane. The break0.493, 0.857, and 2.65. These are obtained
(8-151).
when
1.185,
it
can be
and s = 1 converge to
a point. The root loci for this situation are sketched as shown in Fig. 8-28(c).
When a is greater than 1.185, Eq. (8-151) yields one real root and two complexconjugate roots. Although complex breakaway points do occur quite often in root loci,
we can easily show in the present case that these complex roots do not satisfy the original equation of Eq. (8-149) for any real K. Thus the root loci have only one breakaway
point, as shown in Fig. 8-28(d) for a = 3. The transition between the cases in Fig.
8-28(c) and (d) should be apparent.
shown
8.6
that the
lie
between s
The root
a controller that
is
it is
when
these poles
values.
&
-'
""
ft
I!
XI
ll
o
-C
"3
+
r
(1)
J3
**
fc
3 *
.S
r*i
3
O '4
II
II
<3
&5
& ^
00
90 .
11
11
v^2-
as a;
II
DC
H. "b
425
>t
t
if
-*
3
8
*-
H^
o
II
./
t^r
/^"^
2
*
""
-*
T
7
I
l
8
\
/
1
\
a)
to
2'
v.: \^y
en
1
e
8
1
426
Sec. 8.6
427
In Section 8.5 the root locus diagrams of equations with two variable
parameters are studied by assigning different values to one of the parameters.
In this section the multiparameter problem is investigated through a more
systematic method of embedding. When more than one parameter varies con-
oo to oo, the loci of the root are referred to as the root contours.
be shown that the same conditions and rules of the root loci are still
applicable to the construction of the root contours.
The principle of the root contours can be illustrated by considering the
tinuously from
It will
equation
Q(s)
+KP
K.PAs)
2 (s)
(8-153)
where K and K2 are the variable parameters and Q(s), P^s), and P2 (s) are
polynomials of s. The first step involves the setting of one of the parameters
equal to zero. Let us set K2 equal to zero. Then Eq. (8-153) becomes
t
+ K^is) =
Q(s)
(8-154)
The root
Kx
Z^=
(8-155)
(8-156)
or
The
G,(j)ff,(s)
G^H^s) =
Next,
we
(8-153), with
2,
^^
(8-157)
We
have
or
1
+G
(s)H2 (s)
(8-159)
upon the
poles and
^'^ ^HW)
is
(8 - 160)
G 2 (s)H2 (s)
are identical to
variable parameters.
Example 8-22
where
KA
and
+ K s 2 + K iS + K =0
2
(8-161)
lie
between
and
oo.
428
Chap. 8
As a
first
step
we
let
K =
2
0; Eq. (8-161)
.y
which
is
K^s
becomes
=0
Ki
(8-162)
converted to
(8-163)
The root
loci of
as
shown
in Fig. 8-29(a).
Next, we let
2 vary between zero and infinity while holding Ki at a constant
nonzero value. Dividing both sides of Eq. (8-161) by the terms that do not contain K2
we have
,
K s*
2
(8-161)
+K
K,s
when
(8-165)
t
varies
may
configuration of
G 2 (s)H2 (s)
=
s
KiS
(8-166)
K,
iu
f
AT,
s-plane
K, =
AT,
=0
0.5
(a)
Root contours
for s 3
+ K2 s 2 + K
+K =
t
0,
K2 =
Root Contour
Sec. 8.6
Multiple-Parameter Variation
429
s-plane
K2
/
=0(0.4+/1.74)
K, = 2.56
(b)
K2
The
varies,
zeros of
K =
t
Root contours
for s 3
K2S i + K iS + K
-\
constant.
G 2 (s)H2 (s)
are at s
0, 0;
G^H^s)
which have been found on the contours of Fig. 8-29(a). Thus for fixed A"i
the root contours when K2 varies must all emanate from the root contours of Fig.
8-29(a).
Example
8-23
G{s)H{s)
s{[
The
It is
Ts)(s z
Ts)(s
(8-167)
1)
variable parameters.
2s
K and T as
2
2s
is
2)
written
+K=
(8-168)
430
First,
Chap. 8
we
T equal
shall set
s(s
The
to zero.
2
characteristic equation
+ 2) + K =
K varies are
2s
equation when
this
becomes
(8-169)
zeros of
Gi(s)Hds)
as
shown
A,
+ 2s + ^
rf . 2
s(s
.l.
(8
2)
"
17 )
in Fig. 8-30(a).
fU)
s-plane
/t = o
/ K=\0
+/1
K=
/CO
..
s-plane
K=
*>%_
^K
T= oo
K=
<
]
o"t = ~
k=io
T=
K=4
K=
K=
\\
T=
oo
a:=io
\ T=
\
(a)
Root
figuration of
(b)
4-
2s
2)
+K=
0. (b)
[Ts 2 (s 2
2s
2)]/s(s 2
G 2 (s)H2 (s) =
2s
2)
G^tfaCs)
Ts
(s
s(s 2
when T
K].
+ 2s + 2)
+ 2) + K
(8-171)
2s
Example 8-24
As an example
0,
the points
G(s)H(s), consider
G(s)H(s)
K(l
1)(*
s(s
Ts)
(8-172)
+ 2)
Let us
first
l)(s
2)
K(l
as discussed
+Ts)=0
(8-173)
T=
in
Eq. (8-173)
Root Contour
Sec. 8.6
Multiple-Parameter Variation
431
s-plane
Root contours
Fig. 8-31.
s-plane
/
r=i
o<-r
for s(l
sT)(s 2
^\.
2)
CO
r-K
Root contours
for s(l
4.
7^0^
T'-*
oo
r->-=c
r=0
jT-^-oo
K>
0;
^
MJ
O^T
+K=
v" r =
T^KX>
r->
2)
r-^-oo
/ T=0
'o
i
s-plane
T=0
Fig. 8-32.
2s
X
\
sT){s 2
\
+
2s
Root contours
Fig. 8-33.
+ K = 0; K = 0.5.
2)
+K=
0;
K<
for
,j(l
\
+ sT)(s 2 + 2s
0.5.
yields
*(j
l)(s
2)
+K=
(8-174)
(8-175)
which leads to
s(s
The root
loci
l)(s
2)
of Eq. (8-174) are sketched in Fig. 8-34, based on the pole-zero configu-
ration of
When T varies
infinity,
+ G 2 (s)H2 (s) =
we
+ s(s +
TKs
l)(s
+ 2)+K
(8-177)
432
Chap. 8
/CO
s-plane
K=
K=0
*-AT
-K-
Fig. 8-34.
Root
K=0
K=
-*
\',K
= 6
1)0
2)
+K=
/OJ
0.
/
,*K =20
5-plane
AT
= 20
+H^
iHt-
-3.85
0.425-/2.235
X
K],
K=
G 2 (s)H2 (s) =
K=
20
\\
TKs/[s(s
1)0
+ 2)
20.
The points that correspond to T = on the root contours are at the roots of
+ 1)0 + 2) + K = 0, whose loci are sketched as shown in Fig. 8-34. If we choose
K = 20, the pole-zero configuration of G (s)H2 (s) is shown in Fig. 8-35. The root
contours of Eq. (8-173) for
< T < co are sketched in Fig. 8-36 for three values of
s(s
Sec. 8.6
433
s-plane
\a:=20, r=o
Fig. 8-36.
l)(s
2)
+K+
is
KTs
0.
is,
-3.85
0.425
+ 0.425 =
-1.5
(8-178)
sum
is
is
zero.
434
Cna _
The root contours shown in Fig. 8-36 verify the well-known fact that the derivative
control generally improves the relative stability of the closed-loop system by moving
the characteristic equation roots toward the left in the s-plane. The root contours also
clearly indicate an important characteristic of the derivative control in that the bandwidth of the system is increased. In certain cases the contribution to the increase in
bandwidth by increasing the value of T far exceeds the improvement made on the
relative stability of the system.
bilized for all values of
As shown
However, the
K = 20,
largest
the system
damping
is
sta-
ratio that
8.7
is
we
shall
KP{s)e' T '
s.
may be
written
(8-179)
An
is
1
(8-180)
where
G.fr)^) =
Thus, similar to the development in Section
the following conditions must be
where
=a
= 5] in
8.2, in
met simultaneously:
= j~
IG^Hjjs) = (2k +1)ti +
/G^H^s) = 2kn + (oT
+ joo and k = 0, 1, 2,
(8-181)
gg
- oo < K <
coT
co
K>0
K<0
(8-1 82)
(8-183)
(8-184)
point s
T=
it is recognized that if T
0, given a value of K, there are
only n points in the s-plane that will satisfy either Eq. (8-183) or Eq. (8-184),
for all possible values of k, where n is the highest order of P(s) and Q{s). How-
another standpoint,
ever, for
I?t0,
co,
may
Sec. 8.7
435
be more than n points which satisfy the angular conditions in the s-plane, as k
on all possible integral values. In fact, there are an infinite number of
takes
is
transcendental,
number of roots.
The difficulty with the construction of
is
known
to
have an
infinite
many
is
that
K=
Points
Theorem
The
8-9.
K=
and a
Equation (8-182)
Proof:
= .
is
repeated,
e-^IG^)/^)^^
Thus,
if
The
(8-185)
G^H^s),
or a, which
is
approaches -co.
s,
K co
Points
Theorem
8-10.
The
Proof:
evident.
Number of Branches on
(8-179)
is infinite,
since
is
loci are
real coefficient
but with
infinite
order.
where
is
(8-186)
436
Chap. 8
=
m=
n
number of finite
poles of
number of finite
zeros of
G^H^s)
G^H^s)
Table 8-2
>o
K=
N = even integers
Odd
=
N=
=
N=
=
N=
=
Even
<o
Asymptotes
Odd
Even
odd integers
1, 3, 5,
...
even integers
0,
2, 4,
Asymptotes
N odd integers
2, 4,
odd integers
1, 3, 5,...
0,
K =
= 1, :t3, 5,...
= odd integers
= 1, .b3, 5,...
N = even integers
= 0, 2, 4,
N = even integers
= 0, 2, t4,..
iV
Proof:
Since as s
(8-184).
The property of the root loci of Eq. (8-179) on the real axis is the same as
Theorem 8-7, because on the real axis, co = 0, the angular conditions
of Eqs. (8-183) and (8-184) revert to those of Eqs. (8-9) and (8-10), respectively.
stated in
Intersection of the
branches. However,
we
shall
show
sections nearest the real axis are of interest for stability studies.
Breakaway Points
Theorem
must
satisfy
8-12.
loci
ofEq. (8-179)
Sec. 8.7
dG
(.s)H 1 (s)e-
T'
437
(8-187)
ds
Proof:
is
similar to that of
Theorem
8-8.
<r
Example
is
8-25
on
is
(8
"
188)
+ Ke- T *=0
s
It is
st
= Tr77^T771T
|Gl(Sl)lC5l)l
l*l
where
is,
(8-189)
desired to construct the complete root loci of this equation for a fixed value of T.
we
s,
get
+ s
Kp~ Ts
which
is
(8-190)
CPifj) =
The following
1.
The
K=
From Theorem
points:
47t/r,
approaches
zero as
K< 0:
A"
The K =
shown in Fig.
The root
0,
-co
at co
and
at
-co.
= 0,
2a/T,
a approaches -co
at co
n/T,
K=
oo
at
371/7",
as
approaches
+co
at
=
ft)
co.
Using
nlT,
co as a approaches +oo at co =
0,
.1njT,
K=
8-37.
loci,
loci
The
approaches
47r/r;
of the root
+oo
approaches
37t/r,
at s
K>0: K
4.
K=
we have
K < 0: K approaches
571 IT,
8-9,
8-2,
3.
(8-191)
2.
oo
points,
The notation of
and 0~
is
for the
is
loci are
complementary root
loci.
the
complemen-
<
intersections of the root loci with the jco axis are relatively easy to
determine for
Since
this
simple problem.
G i(s)Hi(s)
0, for
any point
si
on
438
Chap. 8
/CO
s-plane
^K
j4ir/T
K^-oo
0-^K
Pn/T
K~>+
^K
I2v/T
K^
MT
K^+
dzi^.
K^
-oo
-jn/T
->
-J2*IT
a:-*
-oo
0-^K
-miT
K-* +oo
+ <-K
-j4*/T
0-+-K
+
^K
Q-^K
+
+-
Fig. 8-37.
loci
of the equation s
oo
-oo
<*>
-o
Ke~ Ts
0.
/GifoWifri)
and for any point
s^
on the negative j CO
/gi(Ji)tfifri)
Thus, for
axis (co
>
K>
0,
(8-192)
axis,
=4r
(8-193)
loci
on they CO
0),
-~ = (2k +
k
--
= 0, 1,
2,
The values of co
l)7t
+coT
(8-194)
n_ 5n_ 9n_
(8-195)
Sec. 8.7
For
K > 0,
and
<a
= 0,
1, 2,
CO
Similarly, for
439
< 0,
=
~-
and for k
(2k
(8-196)
= -jj, ~2j,
K < 0,
+ coT
l)w
~2j;,
(8-197)
...
-y = 2ta +cor
y=
The crossover
the last
2A:7t
co>0
+ cor
<
co
two equations.
(8-198)
(8-199)
= 0,
1, 2,
G>=f|,2y,^,..5.
into
We have
(8-200)
dG (s)H
i
(s)e-Ts
ds
d^\
ds\
loci are
(8 . 201)
or
Te~ Ts s
e~ T
(8-202)
The
1)
is
(8-203)
at s
IjT.
values of
Eq. (8-188).
<7
shown
loci
of Eq. (8-189)
in Fig. 8-38.
infinite number of roots, and therenumber of branches, from the system analysis standpoint, only the branches that lie between njT < co < n\T are of interest. We shall
refer to these as the primary branches. One reason is that the critical value of K at the
crossover point on this portion of the root loci is equal to 71/27', whereas the critical
K at the next branch at co = 5nj2T 571/2T, which much greater. ThereK = 7C/Tis the critical value for stability. Another reason for labeling the primary
that for any value of K less than the critical value of
branches as the dominant loci
value of
is
is
fore,
is
7t/2r, the
all
440
Chap. 8
s-plane
+ <-K
+ <-K
Complete root
Fig. 8-38.
loci for j
Ke~ Ts
0.
Therefore, the transient response of the system, which has Eq. (8-189) as its characteristic equation, is predominantly controlled by the roots on the primary branches.
Example 8-26
As a
loci
slightly
tem shown
in Fig. 8-39.
G(s)H(s)
The loop
Ke~ T
s(s
The
characteristic equation
is
1)
is
Sec. 8.7
^
9
r(t)
441
cU)
s{s+
l)
Fig. 8-39.
-Ts
in the
feedback
path.
+s+
s2
by using the
1.
rules given in
Theorems
Ke~ Ts
loci,
8-9
through 8-12.
5n /T,
K<0: K
4n/r,
2.
approaches zero as
.
approaches
K>0: K
K<0:
+co
approaches
37t/T,
4a/T,
at co
= 0,
2tc/T,
as
8-10,
co.
Using
+co
at
co
njT,
at co
=0,
K = oo
approaches
at
approaches
A'
co
co
as
a approaches +co
27t/r,
K=
3.
root loci
4.
(AT
The
<
rest
is
0).
s=jf [-(T+2)*/T*
1
For
T=
sec the
-0.382
at
+4]
at
-2.618
loci are
442
Chap. 8
s-plane
^K
+ <-K
Complete root
Fig. 8-40.
loci for s 2
+s+
Ke~ T
0,
T=
1.
loci,
The complete root loci of the system are sketched as shown in Fig. 8-40 for T
1
Notice that from the system analysis standpoint, only the portion of the root loci
that lies between CO
and CO = n is of significance. The closed-loop system of
sec.
=n
Fig. 8-39
is
stable for
<:
K<
way of
in the preceding section point to a systemconstructing the root loci of a closed-loop system with pure time
still
be quite
difficult.
Sec. 8.7
5.
One method
is
Ts
,
443
by a poly-
to approximate e'
Ts
,
as
follows
e
"
[1
(8-205)
(Tsln)Y
Since e' Ts has an infinite number of poles, the approximation is perfect when n
becomes infinite. Figure 8-41 illustrates the effect of the approximation when
the input to the pure time delay
is
(b)
Fig. 8-41.
effect
of approximating a pure
time delay by
If
Eq. (8-205)
is
the primary branches of the root loci will be realized. However, this will be
adequate for the great majority of practical problems, since only the primary
branches will contain the dominant eigenvalues of the system.
Let us approximate the exponential term of Eq. (8-189) by the right side of
2, 3, and
Eq. (8-205). Figure 8-42 illustrates the dominant root loci for n
approxiroot
loci.
The
primary
branch
the
exact
together
with
the
of
4; T
1,
is
to
is,
= 1
+ T2!s
2
Ts
7V +
(8-206)
3!
The
terms of the
Chap. 8
/oo
s-plane
Fig. 8-42.
(1
8.8
Relationship
s/n)"
Between Root
+K=
Loci
loci
of 5
Ke~ s
by those of
0.
In Chapter 7 the polar plot of a transfer function G(s)H(s) is shown to be a mapping of the imaginary axis of the s-plane onto the G(s)H(s) plane. In general,
mapped onto
the function G(s)H(s), such as in the case of a system with pure time delay.
Sec.
Relationship
8.i
G(s)H(s)
Plot /
= + KG ^H i{s) =
445
(8-207)
KGi(s)Hi(s)
KG
(s)Hi(s)
=
=
2kn
mapping of the
the root
loci,
and
fact, for
K>0
K<0
+l)n
(2k
the
(8-209)
real axis
K > 0,
(8-208)
complementary root
for the
loci,
K < 0, the mapping points are on the positive real axis of the j-plane.
It
was shown
in
0Ji
on the imaginary
is
which
is
is
Im
is
shown
GH
in Fig. 8-43.
/CO
s-plane
Mapping from
s to
GH-pXsme
ReGH
The
K/[s(s
a)(s
b)]
interpreted as a
446
Chap. 8
/'
.
Im
GH
G(s)H(s)-p\ane
/ ReGH
Rp (
lGH=2kv^
lGH=(2k+
1)tt
Fig. 8-44.
preted as a
Root locus diagram of G(s)H(s) = Kj[s(s + a)(s + b)] intermapping of the real axis of the G(s)H(s)-p\ane onto the j-plane.
the j-plane.
As shown by curve
Im
when
GH
G(s)H(s)-pUne
/CO
s-plane
root of the
characteristic equation
ReGH
in the s-plane.
constant-damping ratio
lines
Sec. 8.9
447
through the ( l,y"0) point, it means that Eq. (8-207) is satisfied, and the corresponding trajectory in the j-plane passes through a root of the characteristic
equation. Similarly, we may construct root loci that correspond to straight
from the
by
Fig. 8-46.
/KG^H^s) =
loci
now
satisfy the
(2k +1)ti
-6
shown
condition of
K>
(8-210)
0.
G(s)H(s)e je
(8-211)
Im
GH
s-plane
Fig. 8-46.
Root
loci that
G(s)H(s)-p\a.ne. (The
complementary root
quency
8.9
co.
The root
The
characteristic
equation roots of the discrete-data system having the block diagram of Fig. 8-47
448
Chap. 8
must
satisfy
if
(8-212)
if
GH*(s)
the z-plane
is
GH{z)
(8-213)
referred to.
Since
GH*(s)^-L
G(s
n=-o
(8-214)
an infinite series, the poles and zeros of GH*(s) in the j-plane will be
number. This evidently makes the construction of the root loci of Eq.
(8-212) more difficult. However, as an illustrative example of the difference
between the characteristics of the root loci of continuous-data and discretewhich
is
infinite in
data systems,
let
G(s)H(s)
s(s
Using Eq.
(8-214),
(8-215)
1)
we have
GH*(s)
= -L^
(8-216)
+ jncOs)(s + jna> + 1)
and s = jnco where n takes on all integers
(s
for positive
tain
an
infinite
loop system
well
known
is
that the
all
values of
8-48(b) for
T=
The root
loci
con-
K greater than
is
4.32. In contrast,
it is
values of K.
The root
is
Sec. 8.9
.s-plane
X--
449
,/co
s-plane
^K = 4.32
a:-*
:;-
^K
'_JL
/2co s
--)r
K-
3co s
oo<-
K-too
/" s
x-
it
^K = 4.32
oo<-a:
K=
,K =
x,
^-K = 4.32
<~*-K
.
--X
it
,K =
3 co t
'V
^-K = 432
oo^a:
*:->>
/2cos
--*
K=
it
5co.
-/-
^K = 4.32
<-a:
-;3cos
(b)
K/[s(s
for
1)],
K^-<*>
(a)
the
discrete-data
T=
sec.
K-k*>
number of root
is,
in
loci is
and the same rules of construction for continuous-data systems are directly
applicable.
As an
illustrative
let
and poles at z
1 and z
0.368. The root loci for
the closed-loop system are constructed based on the pole-zero configuration of
Eq. (8-217) and are shown in Fig. 8-49. Notice that when the value of AT exceeds
one of the roots of the characteristic equation moves outside the unit
4.32,
450
Chap. 8
Imz
z-plane
f-path (f =
50%)
Rez
Unit
circle
Fig. 8-49.
K/[s(s
+
,
1)],
Im
T=
sec.
z-plane
Unit
circle
Rez
Fig. 8-50.
Root
if
Kl[s(s
1)],
T=
5 sec.
is
changed to
T=
5 sec, the z-
GH(z)
0.993Kz
=
(z
l)(z
0.0067)
(8-218)
Sec. 8.9
The root
when T
K for
marginal K of 4.32 for T =
stability for
1
loci for
T=
T=
system
sec, the
451
be noted
5 sec takes the form of
drawn as shown
is
5 sec is 2.02 as
compared
to the
sec.
per cent.
GM GH{z) _ K[(T
e~ T )z
(z
Te~ T
l)(z
+ 1
e~ T ]
(8-219)
e-*)
The root loci of the system with sample-and-hold for T = 1 sec and T = 5
shown in Fig. 8-5 1(a) and (b), respectively. In this case the marginal
sec are
/Iraz
z-plane
Rez
(a)
Fig. 8-51.
Root
loci for
loci for
T=
sec
ple-and-hold. G(s)H(s)
Root
Root
T=
5 sec.
= K/ls(s +
1)].
(a)
Root
loci for
T=
sec. (b)
452
Chap. 8
z-plaue
(b)
Root
loci for
T=
5 sec
K is
2.3 for
T=
sec
and 0.66
for
T=
5 sec.
This
illus-
REFERENCES
General Subjects
1.
W.
W. R. Evans,
4.
Chu and
V. C.
M. Yeh, "Study
J.
Trans.
New
ASMS,
G. Truxal, Automatic Feedback Systems Synthesis, McGraw-Hill Book ComNew York, 1955.
pany,
6.
K.
Chap. 8
7.
P.
References /
Mosner,
J.
Nonminimum Phase
Analysis of
"A
453
Systems,"
IRE
B.
J.
IRE Trans.
Automatic Control,
J.
D.
S.
10.
Trans. Automatic
Control, Vol.
H. M. Power, "Root Loci Having a Total of Four Poles and Zeros," IEEE
Trans. Automatic Control, Vol. AC-16, pp. 484-486, Oct. 1971.
C. C.
29-104, 1954.
12.
13.
J.
14.
E. Gibson, "Build a
Dynamic Root-Locus
IRE
15.
16.
C. A. Stapleton,
Cong., Vol.
1,
Control, Vol.
17.
IRE
Trans. Automatic
18.
C. F. Chen,
July 1965.
19.
20.
AC-U,
Trans. Automatic
R.
J.
23.
J.
1966.
"Some
Invariances of the
454
Chap 8
24.
25.
Plotting,"
Root
27.
28.
29.
D, Vol.
86,
IRE
Trans. Automatic
ASME,
J.
Basic
Mar. 1964.
C. S. Chang, "An Analytical Method for Obtaining the Root Locus with Positive
and Negative Gain," IEEE Trans. Automatic Control, Vol. AC-10, pp. 92-94,
Jan. 1965.
30.
31.
B. P.
J.
Control, Vol.
J.
D.
S.
IEEE
Internat.
D.
J.
May
33.
Doda, "The
Digital
Eng.,
1958.
Dec. 1968.
34.
36.
Trans.
F.
Design," Trans.
1956.
37.
J.
IRE Natl.
38.
2,
39.
Problems
Chap. 8
455
A. Katayama, "Semi-graphical Design of Servomechanisms Using Inverse RootLocus," /. IEE, Vol. 80, pp. 1140-1149, Aug. 1960.
40.
41.
42.
43.
J.
IRE
W. L. McDaniel, Jr., "A Generalized Root Locus FollowIEEE Trans. Automatic Control, Vol. AC-15, pp. 483-485, Aug.
R. Mitchell and
ing Technique,"
1970.
Root
Sensitivity
44.
J.
45.
Trans.
AIEE Appl.
46.
IRE
Nov.
1963.
48.
B. C.
Prentice-
Hall, Inc.,
49.
B. C.
A, Champaign,
111.
2277, Station
1974.
M.
J.
anisms,"
IRE
AC-4, No.
3,
1959.
PROBLEMS
8.1.
Sketch the root locus diagram for each of the following feedback control systems. In each case determine everything about the locus of roots for oo <
K<
co and sketch the root loci. Indicate on each locus the starting point, the
ending point, and the direction of increasing value of K. The poles and zeros
456
chap
(a) Poles at 0,
(b) Poles at 0, 0,
Poles at
(c)
-2 +j2, and -2
-j2; zero
at
-3.
(d) Poles at 0,
Poles at
(e)
8.2.
(f)
Poles at
(g)
The open-loop
-12;
-4
zeros at
and -8.
K '
s(s 2
2s
2)(s
5)(s
is
given by
6)
8.3.
K
w\- s\l
+ - 2s)( + 0-0255)
+ 0.001.0(1 + 0.0050
< K < co) root locus diagram for the system. Indi<1
C(
on the j CO
8.4.
K and co
at these points.
G(i)
=
s(\
(c)
The
G(S)
(a)
+ 0.020(1 + 0.010
Sketch the root locus diagram of the system (0 < K < co).
Determine the marginal value of K that will cause instability.
Determine the value of K when the system is critically damped.
(a)
(b)
8.5.
axis,
of
= s\s +
2)(s
and
5)
"W =
'
Sketch the root locus diagram for the system. Indicate the crossing points
of the loci on they axis and the corresponding values of
and co at these
K only).
The
H(s)
is
now changed
to
of K. Investigate the effect on the root locus diagram due to this change in
H{s).
8.6.
The
3.s
+(K +
2)s
given by
+ 10K =
is
system.
For the following loop transfer function, sketch the root locus diagram as a
T (T varies from to 00). Determine the value of T so that the
damping ratio of the complex roots of the characteristic equation is 0.2.
function of
ww =
G(s)H(s)
s(l
1000(1 + Ts)
+ 0.10(1 + 0.0010
Problems
Chap. 8
8.8.
457
For the following loop transfer function, sketch the root locus diagram as a
function of T. Determine the value of Tso that the damping ratio of the complex
roots of the characteristic equation
G(s)H(s)
is
0.2.
30
j(1
8.9.
+ a2j)(1 +
Q U)(1
Ts)
C2
+ o-
-VW\r
-VW\r-
-o +
C,
-oFigure P8-9.
(a)
Sketch the root locus diagrams of the zeros and poles of E1 jE l as a function
to oo).
(C, varies from
of
(b) Sketch the root locus diagrams of the zeros and poles of
tion of C 2
2 /i
as a func-
8.10.
The open-loop
is
given by
G(s)
s(s 2
+ 4s + 4)
It is
K (0 < K <
oo).
C(s)
R(s)
(1
2
)
Determine the open-loop transfer function G(s) of the system. Assume that the
system has unity feedback.
8.12.
as 1
Ks
+K=
oo < K <
co and
(a)
(b)
applicable.
8.13.
The open-loop transfer function of a control system with unity feedback is given
by
G(s)
K{s
s 2 (s
+
+
a)
1)
458
Chap. 8
Determine the values of a so that the root locus diagram will have zero, one,
and two breakaway points, respectively, not counting the one at s = 0. Sketch
the root loci for
8.14.
co < K <
co for
all
three cases.
in Fig. P8-14,
is:
s(l
+ 0.2s)
r(t)
G(s)
z.o.h.
c(t)
Figure P8-14.
(a)
T=
0.1 sec
system (0
T=
and
sec.
<K<
8.15.
Repeat part
(a)
when
loci (0
Kz 2
\.5Kz
<K<
K for
oo < K <
stability.
-(K+\)=0
r(t)
^
j-/
^
T=
0.
Ke -0.is
z.o.h.
1
s(s
sec
Figure P8-16.
)(s
c(t)
3
9
Frequency-Domain Analysis
of Control
9.1
Systems
Introduction
was pointed out earlier that in practice the performance of a feedback control
is more preferably measured by its time-domain response characteristics.
This is in contrast to the analysis and design of systems in the communication
field, where the frequency response is of more importance, since in this case
most of the signals to be processed are either sinusoidal or periodic in nature.
However, analytically, the time response of a control system is usually difficult
It
system
there are no unified ways of arriving at a designed system given the time-domain
specifications,
On
rise time,
is
based on the relationships that exist between the time-domain and the frequency-domain properties. Therefore, we may consider that the main purpose
of conducting control systems analysis and design in the frequency domain is
merely to use the techniques as a convenient vehicle toward the same objectives
as with time-domain methods.
The
starting point in
frequency-domain analysis
is
We shall first discuss transfer function relations based on, first, the state variable
representation and, second, the classical approach.
system
is
459
460
Frequency-Domain Analysis
relation
is
of Control
Systems
CO)
where CO)
Chap. 9
is
a9
function matrix
x
is
vector
[I
and R0)
is
a q
x p
Under the
[I
G(j)H(*)]" G(*)
(9-2)
represents the
we
set s
+ GUaWVmfi-iGUm)
[I
(9-3)
defined as
is
MM =
1
vector.
is
matrix.
M(./ca)
where
ap X
(9-1)
defined as
MO) =
which
G(j)H(j)]- G(j)R(j)
row and j
1X J\JU')
(9-4)
all
other inputs=0
column of M(jco).
the
State-Variable Representation
is
c(0
=
=
+ Bu(r)
Dx(?) + Eu(0
Ax(f)
(9-5)
(9-6)
is
described by
= r(0 - Hc(0
(9-7)
where
= nxl
UW = P X
c(0 = 9 X
r 0) =
x
x(?)
/>
state vector
control vector
output vector
input vector
A, B, D, and
p X g
CO)
The open-loop
[DOI
- A)"B + E]U0)
GO)
The closed-loop
is
is
the
(9-8)
defined as
= DOI - A)-B + E
is
(9-9)
[I
G0)H]" G0)R0)
(9-10)
MO) =
[I
G0)H]-'G0)
(9-11)
CO)
Thus
It
H in Eq.
(9-1 1)
Introduction / 461
Sec. 9.1
In general, the elements of the transfer function matrices are rational funcit is proved that if the system is completely controllable
tions of s. In Chapter 4
and observable, there will be no pole-zero cancellations in the transfer functions. Under this condition the poles of the transfer function will also be the
eigenvalues of the system.
The
domain discussed
in the following
sections are conducted with the single-variable notation. Because linear systems
satisfy the principle
all
be applied
to multivariable systems.
transfer function
is
written
^>=)= r+wm
Under
M( fm)
MUC0)
we
set s
(9 " 12)
C^^ G Jc)
~ RUa>)
~ + GUcoWJco)
U)
(9-13)
{
*
The
function of co,
is
a complex
Re
[M(jco)]
+ j Im [M(jco)]
(9-14)
M(jco)
(9-15)
where
G(jco)
1
(9-16)
+ GUco)HUa>)
and
<l>m(co)
Gjjco)
+ G(j(o)HUco)
= IG{jg>) - /l + GUa>)H{j<o)
1
is
now
in the frequency
For
may be
is
(9-17)
The
may be
regarded as the
significance of M(a>) to a
an electronic
is
amplifier.
must have a flat gain for all frequencies. Of course, realistically, the design
criterion becomes that of having a flat gain in the audio frequency range. In
control systems the ideal design criterion
is
similar. If
unity for
all
frequencies.
all
it is
desirable to keep
frequencies, M(ja>)
(9-13)
it is
must be
462
Frequency-Domain Analysis
of Control
Systems
Chap. 9
can be unity only when G(joi) is infinite, while H(jco) is finite and nonzero. An
infinite magnitude for G(jco) is, of course, impossible to achieve in practice, nor
would it be desirable, since most control systems become unstable when its
loop gain becomes very high. Furthermore, all control systems are subjected to
noise. Thus, in addition to responding to the input signal, the system should be
and suppress noise and unwanted signals. This means that the
frequency response of a control system should have a cutoff characteristic in
able to reject
an ideal low-pass filter, which is impossible to realize physTypical gain and phase characteristics of a feedback control system are
shown in Fig. 9-2. The fact is that the great majority of control systems have
characteristics of
ically.
filter,
increases.
M(a>)
*m(")
Fig. 9-1.
filter.
M(u)
9.2
and phase
Frequency-Domain Characteristics
If a control system is to
we need a
set
Frequency-Domain Characteristics
Sec. 9.2
Peak resonance
value of M(co) that
p.
is
as the
p is defined
463
maximum
gives
is
may be necessary
from
noise.
Sometimes
it
However, in general, a steep cutoff characteristic may be accomcorresponds to a system with a low stability margin.
p which
The performance criteria defined above for the frequency-domain analysis
frequencies.
panied by a large
are illustrated
criteria that
may be used
as
shown
in Fig. 9-3.
in the ensuing
M(cj)
0.707
464/
9.3
p . CO,,
Chap. 9
We may
= C Jco) =
M(jco)
+ Aeo/aOC -
by
(a/a*) 2
letting u
cojco n
becomes
M(yM)
are
= M(M) =
[(1
m2)2
(9-20)
(2Cm)2]1/2
and
MM =
0()
-tan-
'
^Lj
(9-21)
^^ =
-i-[(l
u 2) 2
- 4 +
8C
(9-22)
from which
- 4m +
4m 3
SuC 2
(9-23)
(9-23) are
u,
u,
= J\-
(9-24)
and
The
2 2
(9-25)
solution in Eq. (9-24) merely indicates that the slope of the M(co) versus co
is zero at co
0; it is not a true maximum. The solution of Eq. (9-25)
curve
Since frequency
= co.Vl - 2
(9-26)
>
2
a real quantity, Eq. (9-26) is valid only for 1
2 or
<; 0.707. This means simply that for all values of greater than 0.707, the
solution of (o p
becomes the valid one, and p
1
is
MB
Sec. 9.3
a>,
and
M =
important to note that
M is
we
465
get
2CV1
It is
simplifying,
(9-27)
co p is
a function
of and a>.
It should be noted that information on the magnitude and phase of M(jco)
of Eq. (9-18) may readily be derived from the Bode plot of Eq. (A-51), Figs.
A-10 and A-12. In other words, Fig. A-12 is an exact representation of Eq.
(9-21). The magnitude of M(jco), however, may be represented in decibels versus
frequency, which is the Bode plot in Fig. A-10, or it may be plotted in absolute
magnitude, M(co), versus co. Figure 9-4 illustrates the plots of M(u) of Eq. (9-20)
u = to/co
Fig. 9-4. Magnification-versus-normalized frequency of a second-order
466
Chap. 9
Damping
Fig.
9- 5.
1/(2{V1
p -v ersus-damping
ratio
ratio f
for a
second-order system,
Mp
2 ).
Notice that
if
when
increase
When
u
and 9-6
co p /co
As
system
by Eq.
decreases, as indicated
and
0, co p
co
become
and f
would
(9-26).
between
and
and
BW,
of a
respectively.
the fre-
co p
(<y)
down from
the
we
cent of
zero-frequency value, or 3 d B
its
have
M(u)
0.5
Damping
0.707
[(i
1.0
[(i
*/\
2 2
0.707
(2t;uyy
(9-28)
Thus
ratio J
u*y
Fig. 9-6.
system, u p
u*y
(2{k) 2
V4C -
(9-29)
w2
(l
2C
4 2
(9-30)
In the last expression the plus sign should be chosen, since u must be a positive
real quantity for any f. Therefore, from Eq. (9-30), the bandwidth of the
second-order system
is
determined as
BW = a>.[(l - 2C + V4C - 4 +
4
2]>
(9-31)
Sec. 9.4
Effects of
Adding
467
2.0
0.6
0.4
Damping
Fig. 9-7.
ratio f
2.
3.
The maximum overshoot of the unit step response in the time domain
depends upon only [Eq. (6-96)].
The resonance peak of the closed-loop frequency response M
depends upon only [Eq. (9-27)].
The rise time increases with , and the bandwidth decreases with
the increase of
and
Fig.
9-7.
9.4
The relationships
domain responses
more parameters
When
BW
Fig. 9-8.
function G(s)
(1
Ts)l[s(s
transfer
1.414)].
1.2
1
G(s)'
1.0
X
0.8
(1
Ts)
s(s+ 1.414)
vv
^\5
-~L_
0.707
0.6
3
5=
0.4
0.2
^^4
^^
r=o.i
^.
^2rr^"
T=0.0l
i
1
co
rad/sec
Fig. 9-9. Magnification curves for a second-order system with an openloop transfer function G(s).
468
Effects of
Sec. 9.4
Adding
a Zero to the
Open-Loop
Transfer Function /
469
cj rad/sec
(1
Ts)/[s(s
an open-
0.4)].
The closed-loop
may be
is
more realistic
considered as that
ml
G(s)
s{s
at s
1/Tto
<?(*)
(9-32)
2co)
2
ft) (l
s(s
Ts)
(9-33)
2c0)
and
is
repeated here:
M(s)
In principle,
_Qs)
p , cop ,
+ Ts)
+ TcoDs + ml
coljl
R(s)
and
BW
(2c
all
(9-34)
470
Frequency-Domain Analysis
1.2
of Control
Systems
Chap. 9
r=o
0.8
7^2-^^^=sa^^
0.6
0.4
/T=0A /y\
r=
Iff
o.oi
0.2
same
(1
Ts)/[s(s
an open-loop
1.414)].
parameters in
> ,
BW =
(^ + {^ + 46,:)'
(9-35)
where
b
4C
co
4Cco T
2col
o>
(9-36)
BW
and 9-10 give the plots for M(co) for the closed-loop
system that has the G(s) of Eq. (9-33) as its open-loop transfer function; co
1,
n
Tis given various values, and
0.707 and
0.2, respectively. These curves
show
is
Sec. 9.5
Adding
Effects of
a Pole to the
Open-Loop
"i
r=
o.2
0.6
Time (second)
Fig. 9-12. Unit step responses of a second-order system with an open-loop
transfer function G{s)
(1
Ts)j[s(s
0.4)].
T in
which the
BW
decreased by the addition of the zero to G(s). Figures 9-11 and 9-12 show the
However, as
which is at s
= l/T, moves very close to the origin, causing the system to have
a large time constant. Thus Fig. 9-11 illustrates the situation that the
is
causes the time response to drag out in reaching the final steady
9.5
time
state.
rise
but the large time constant of the zero near the origin of the s-plane
fast
transfer function
is
arrived at
by adding
(1
(9-32):
G(s)
= s(s
co
2Cco)(l
(9-37)
Ts)
472
3.0
Chap. 9
"i
The
T 2 co s
that the
1.414)(1
an open-
Ts)].
shown
\j[s(s
(1
We may
BW
4
co
its
is
2
is
r 2 )co* +
(4
a>
2<a
4cw
2> - co =
4
(9-38)
is
now
it
can be
unstable for a certain set of system parameters. However, it can be easily shown
by use of the Routh-Hurwitz criterion that for co
1 and
0.707, the system
is
co
is
effect
also increased.
When T becomes
function
is
to
make
but
show
slightly increased
is
Therefore,
we may
T,T=\
and
T=
5,
the
Sec. 9.6
Relative Stability
Gain
473
Mp
1.4 -
1.0
T=0.s//
0.6
-**^
14
12
10
16
Time (seconds)
an open-loop
\\\s(s
1.414)(1
Ts)].
rise
time increases with the decrease of the bandwidth, and the larger values of
peak overshoots
is
arbitrarily small
>
(co) is infinite,
at IG(jco)
However,
and the peak over-
but
if
stable.
is
When
the
number.
The
general cases.
9.6
Mp
We have demonstrated in the last three sections the general relationship between
the resonance peak
use
The
latter is often
termed
relative stability. In
many
situations
we may
Another
474
Chap. 9
way of measuring
G(/co)//0'cj)-plane
Im
GH
RtGH
K=K
(a)
G(/co)//(;co)-plane
I /
Im
GH
(b)
quency responses.
among Nyquist
plots,
step responses,
and
fre-
Sec. 9.6
C(/'cjM/w)-plane
i I
(c)
475
GH
Im
G(/co)#(/a))-plane
Mp
Im
GH
c(/)l
ReGH
(d)
Unstable system.
among Nyquist
and
frequency responses.
system are shown in Fig. 9-15 for four different values of loop gain K. Let us
shown in Fig. 9- 15(a), in which the loop gain K is low, so the
Nyquist plot of G(s)H(s) intersects the negative real axis at a point (the phase-
response
is
point
shown
away from
the
the system
is still
stable,
is
1 ,
jO)
476
shoot,
and
Chap. 9
is
larger.
stant-amplitude oscillation, as
If AT
is
increased
is
still
shown by
further, the
unstable with
Nyquist plot
becomes
unbounded response,
as
infinite.
( l,/0)
shown
point,
In this case the magnitude curve M(co) ceases to have any significance, and the
only symptom of instability from the closed-loop frequency response is that the
phase curve now has a positive slope at the resonant frequency.
Gain Margin
To
give a quantitative
between the
relative distance
a quantity that
is
called the
gain margin.
Specifically, the gain
margin
( l,yO)
is
G(joo)H(j(o) at co
co c is
is
denoted by
co c ,
|.
Im
GH
G(/co)//(/co) -plane
Phase crossover
CO
CO
ReGH
margin
Relative Stability
Sec. 9.6
Gain
its
Mp
477
is
defined as
gain margin
On
( 1, y'0)
dB.
the
G.M.
On
is
20 log,
it is
^^j,
dB
(9-39)
if
becomes
denned by Eq.
(9-39)
is infinite
in decibels.
When the
G(jco)H(jco) plot goes through the ( 1 jO) point, the gain margin
dB, which implies that the loop gain can no longer be increased as the
system is already on the margin of instability. When the G(jco)H(jco) plot does
not intersect the negative real axis at any finite nonzero frequency, and the
Nyquist stability criterion indicates that the ( 1 jO) point must not be enclosed
,
is
is infinite
in decibels; this
means
that,
infinity before
instability occurs.
When
the
( 1
jO) point
is
magnitude of G{jooc)H(jco c) is greater than unity, and the gain margin as given
by Eq. (9-39) in decibels is negative. In the general case, when the above mentioned condition implies an unstable system, the gain margin is negative in
decibels. It was pointed out in Chapter 7 that if G(s)H(s) has poles or zeros in
the right half of the j-plane, the ( 1, jO) point may have to be encircled by the
G(jco)H(jco) plot in order for the closed-loop system to be stable. Under this
condition, a stable system yields a negative gain margin. In practice, we must
first determine the stability of the system (i.e., stable or unstable), and then the
magnitude of the gain margin is evaluated. Once the stability or instability
condition is ascertained, the magnitude of the gain margin simply denotes the
margin of stability or instability, and the sign of the gain margin becomes
insignificant.
Phase Margin
is merely one of many ways of representing the relative
of a feedback control system. In principle, a system with a large gain
margin should be relatively more stable than one that has a smaller gain margin.
Unfortunately, gain margin alone does not sufficiently indicate the relative
stability
stability
Fig.
of
all
systems, especially
if
variable.
478
Chap. 9
i /
Im
GH
G(/cj)//(/to) -plane
*-
Fig. 9-17.
ferent
ReGH
amount of relative
dif-
stability.
corresponds to a more stable system than locus B. The reason is that with any
change in a system parameter (or parameters) other than the loop gain, it is
easier for locus B to pass through or even enclose the ( 1, jO) point. Furthermore, system B has a much larger
p than system A.
Phase margin
is
plot must be rotated about the origin in order that the gain-crossover point on the
locus passes through the
(/,/))
the angle between the phasor that passes through the gain-crossover point and
the negative real axis of the G(jco)H(jco)-plane. In contrast to the gain margin,
which
gives a measure of the effect of the loop gain on the stability of the closedloop system, the phase margin indicates the effect on stability due to changes of
system parameters, which theoretically alter the phase of G(joS)H(jco) only.
The
analytical procedure of
first
phase margin
where
co g
180
g)
( 1
;0) point
(9-40)
is
encircled
Sec.
Relative Stability
9.1
Gain
Mp
479
i/ImGff
G(/'co)//(/a>) -plane
*-
Fig. 9-18.
ReGH
in the GO'ct>)//(y'co)-plane.
by the G(jco)H(j(o) plot, the gain-crossover point would be found in the second
quadrant of the G(y'co)//(jca)-plane, and Eq. (9-40) would give a negative phase
margin.
Although the formulas for the gain margin and the phase margin are
it is more convenient to evaluate these quantities graphically from the Bode plot or the magnitude-versus-phase plot. As an
illustrative example, consider that the open-loop transfer function of a control
system with unity feedback is given by
simple to understand, in practice
G(s)
10
=
s(l
The Bode
tion of
0.02j)(1
(9-41)
0.2s)
is shown in Fig. 9-19. Using the asymptotic approximathe gain-crossover and the phase-crossover points are deter-
plot of G(j(o)
G(jco)
|,
480
Frequency-Domain Analysis
of Control
Systems
Chap. 9
20
Gain crossover
.^^
1
^v^
Gain margin
^>^- 15dB
20
40
\.
60
^^
80
^^
00
in
10
1
50
.,
w rad/sec
100
1000
100
1000
i
i
-90
Phase margin
^-s^f
180
Phase crossover
T
270
10
50
16
co rad/sec
Fig. 9-19.
Bode
plot of G(s)
10/[j(l
0.2s)(l
0.02s)].
corresponds to the Nyquist plot of G(jco) passing through the ( 1, jO) point,
and the system becomes marginally unstable. Therefore, from the definition of
the gain margin, the gain margin of the system
To
we note
is
15 dB.
is
it
will pass
0.M.
180
125
55
this case,
(9-42)
In general, the procedure of determining the gain margin and the phase
margin from the Bode plot may be outlined as follows
1.
is
measured
G.M.
-|GCA )#(M)l
e
dB
(9-43)
Sec. 9.6
\ 5
v-
^r
20
481
Gain crossover
Gain
margin
u=l7
Phase
margin
^^
yr*
XQ
jT
16
>^30
/ cj rad/sec
<^40
o
40
A60
fioo
-
60
T300
80
OO
00
1
-270
225
180
90
135
Phase (degrees)
Fig. 9-20.
10/[i(l
0.2s)(l
0.02s)].
2.
is
measured
0.M.
180
IG(jcQ g )H(jg} g )
(9-44)
versus-phase plot. For the transfer function of Eq. (9-41), the magnitudeversus-phase plot is shown in Fig. 9-20, which is constructed by use of the data
Fig. 9-19.
is
On
is where the locus intersects the 0-dB axis. Therefore, the gain
simply the distance in decibels measured from the phase crossover to
gain crossover
margin
is
CO
o
"3,
a
D.
4>
a3
fa-
t/i
.S
so
as
fa
<s
83
J3
^c
IS
J--
2?
TO
>
*o '5b
"8
.
e
cS
"o
1-.
cs
cs
c
CS
BO
00
.
_2
.a
&
482
c
%
g
"5
483
Sec. 9.7
- 180
Phase
(c)
Magnitude-versus-phase plot.
dB and
80,
in these
domains.
to the
is
the horizontal
critical point.
Bode
plot,
9.7
Relative Stability
system and the slope of the magnitude curve of the Bode plot of G(jco)H(ja>) at
if the loop gain
of the system
is
may
be
moved
to the region in
;
On
tem
and
if
is
the gain
crossover occurs in the region where the slope of the magnitude curve
The example
is
60
above
is a
dB/decade, the system will definitely be unstable.
monotonically
decreases
as
magnitude
curve
simple one, since the slope of the
co increases.
cited
484
Frequency-Domain Analysis
of Control
Systems
Chap. 9
-20
-40
40 dB/decade
V~ 60 dB/decade
1000
1000
Fig. 9-22.
(1
Bode
plot of
GO)
[K(l
0.2s)(l
0.025s)]/[s 3 (l
0.01s)
0.005.S)].
yj
The Bode
frequency
is
is 1
plot of G(s)
rad/sec,
~~
is
shown
is
K=
negative.
1
The gain-crossover
The closed-loop system
points: one at co
(9-45)
(l
wo phase
12C rad/sec.
crossover
The phase
characteristics
lies in this
gain crossover.
As observed from
Fig. 9-22, at
Constant
Sec. 9.8
M Loci
in
the G(/)-Plane /
485
40 dB/decade,
the system
is
about half of these regions, but even then the resultant phase margin
is small. However, if the gain crossover falls in the region in which the magnitude
curve has a slope of 20 dB/decade, the system is stable.
The correlation between the slope of the magnitude curve of the Bode plot
over
falls in
and the
relative stability
9.8
Constant
M Loci
in
the G(/CO)-Plane
In previous sections
it
was shown
peak
of the closed-loop
Bode
plot
not the usual case, as the open-loop transfer function G(s) is normally
given. For the purpose of analysis we can always obtain the magnification curve
by digital computation on a computer. However, our motivation is to be able
this is
M
specified M
to predict
from the
plots of G(jco),
state,
G(s)
G(jai)
^-
M(s)
G(jco)
=
=
\MUco)\
Re
G^
(9-46)
written
is
+ j Im G(jco)
G(jco)
(9-47)
+jy
Then
M(fi))
G(jco)
1
G(jco)
(9-48)
Jx*
V(i
For
simplicity, let
+ x? +
x) 2
= V* +
2
(9-49)
[(l
+ xY +
2
]
x2
(9-50)
486
Frequency-Domain Analysis
of Control
Rearranging
this
term
x2
which
+
+ yv 2
is finally
M )x + - M )y ~ 2M x = M
2
(1
is
[M 2 /(l -
Chap. 9
equation yields
(1
This equation
Systems
on both
)]
We
sides.
(9-51)
have
M1
2Afl
(1
M1
M2
+ d^ip)V
i-m 2Xx + ({t=W )V = r=m (
,o
9 -52)
simplified to
/(l
|.
=~j
(9-54)
When
takes on different values, Eq. (9-53) describes in the G0'co)-plane
a family of circles that are called the constant
loci or the constant
circles.
The coordinates of the centers and the radii of the constant
loci for various
values of
Mare given
in Table 9-1,
Table 9-1
Center
Constant
loci are
M
M
0.3
0.01
0.33
0.33
0.67
0.7
0.96
1.37
1.0
oo
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.5
3.0
4.0
5.0
6.0
-5.76
-3.27
-2.45
-2.04
-1.80
-1.64
-1.53
-1.46
-1.38
-1.33
-1.19
-1.13
-1.07
-1.04
-1.03
shown
in Fig. 9-23.
Circles
0.5
1.1
CO
5.24
2.73
1.88
1.46
1.20
1.03
0.90
0.80
0.73
0.67
0.48
0.38
0.27
0.21
0.17
2
\
Constant
Sec. 9.8
487
Im G
G(/co)-plane
M = 0.833
ReG
M becomes
infinite,
critical point,
( l,y'0).
M
and the
M=
to the
of the M = locus correspond to values of M greater than
The
than
are for M
of the M =
and those to the
give
Graphically, the intersections of the G(jco) plot and the constant M
desired to
the value of M at the frequency denoted on the G(jco) curve.
than a certain value, the G(jco) curve must not
keep the value of M
the same time must not enclose
any point, and
the corresponding M
with the smallest radius that tangent
point. The constant M
the (
read
to the G(jco) curve gives the value of M and the resonant frequency
M
left
right
1,
real axis.
line
less
line
1.
loci
If
it is
intersect
p less
at
circle at
circle
1, j'0)
is
co p is
p,
on the
G(ja>) curve.
Figure 9-24(a) illustrates the Nyquist plot of G(jco) for a unity feedback
loci. For a given loop gain
control system, together with several constant
the constant
loci give
curve
and
the
between
the
intersects
G(jco)
u
K= K
on the M(co)-versus-co
curve.
is
488
Frequency-Domain Analysis
of Control
Systems
Chap. 9
1/
Im G
G(/'cj) -plane
M=M
M >M
2
ReG
curve,
shown
to be co p2 ,
When
K is increased
to
M=
9-24(b).
'
Sec. 9.9
9.9
in
in
the G(/)-Plane
489
the G(/(0)-Plane
The
is
loci
written as
<t>
m (co)
/Mijco)
we have
Let
tan"
(X) -
tan"
tan
(f>
'
(9-55)
(j-f^)
and
letting
</>
(1/4)
(^6)
j,
+ y*-- =
(9-57)
^y --l+-An=^ + Ai
2
(9-58)
regrouped to give
is
(*
+ ) +
{y
(1/2,
-2^ = 4 + 4^
this
1/2AT)-
<9
The
centers
and the
radii
iy /2
of the constant
and the
Table 9-2
180 n
0,1,2...
-90
-60
-45
-30
-15
AT
tan
m
rm
loci are
Constant
Center x
<>
shown
in Fig. 9-25.
N Circles
= -^-,
2"y =
oo
-1.732
-1.000
-0.577
-0.268
59 >
by
(*&r
are tabulated in Table 9-2,
The
(N 2
m (co),
<f>
xi
When
W- ^ + S +
N=
which
jtt
2N
Radius
= J\ AN 2
0.500
-0.289
-0.500
-0.866
-1.866
OO
15
0.268
1.866
0.577
0.707
1.000
1.931
OO
1.931
30
0.577
0.866
1.000
45
60
1.000
0.500
0.707
1.732
0.289
90
OO
0.577
0.500
490
Frequency-Domain Analysis
of Control
Systems
Chap. 9
,,/ImG
<t>=
15 (- 165)
G(/to)-plane
*-
9.10
Constant
Plane
M and N Loci
in
ReG
the Magnitude-Versus-Phase
shown
when
is
Constant
Sec. 9.10
M and N Loci
in
.,/Im
M=
491
G0'co)-plane
*-ReG
20 log 10 C
Phase
(b)
Fig. 9-26. (a)
Constant
Nichols chart in
addition, the
is
Bode
shifted
made to G(jco)
The constant
tions
in the
M and
transferred
to
the
how
this
is
zeros.
coordinates
without
may
be
difficulty.
circle
may
vertically
magnitude-versus-phase
plane
up or down
Phase
Fig. 9-27. Nichols chart (for phase
from -180
to 0).
tude-versus-phase plane.
The
critical point,
dB and 180
( 1,/)),
in the
loci in the
magni-
dB
12
270
- 260 - 240
140
220'
120
- 100 -90
Phase
Fig. 9-28. Nichols chart (for phase
from -270
to -90).
is
the phase angle that extends from 180 to 0. The chart that corresponds to
the phase from 360 to 180 is a mirror image of that in Fig. 9-27 with
493
494
Frequency-Domain Analysis
of Control
180
respect to the
Systems
axis.
Chap. 9
of phase for
M on these constant M
bandwidth, the
3-dB locus
loci are
many
is
shown
for
90
to
given in decibels.
To determine
should be used.
The following example will illustrate the relationships among the analysis
methods using the Bode plot, the magnitude- versus-phase plot, and the Nichols
chart.
Example
9-1
When
is
is
repeated here
25QA
G{s)
The Bode
is
s(s 2
drawn
as
50.5s
shown
(9-61)
1725)
200.
The gain
dB
41.5
100
463
co rad/sec
Fig. 9-29.
Bode
Example
9-1
margin and the phase margin are determined from the Bode plot to be 5 dB and 40,
The data on the magnitude and phase of G(jco) are transferred to the
respectively.
magnitude-versus-phase plot of Fig. 9-30. From Fig. 9-30, the peak resonance
p is
found to be approximately 5 dB (or 1.78), the resonant frequency is 40 rad/sec, the
bandwidth of the system is 48 rad/sec, and the results for the gain and phase margins
are as given above.
When
the
motor inductance
is
dB
Constant
Sec. 9.10
and
Loci
in
495
-4
-180
-160
-140
-120
-100-90
Phase (degrees)
Fig. 9-30. Magnitude-versus-phase plots for the system in
Example
9-1
With
A =
200,
and
500s
(9-62)
17270)
G{s)
Bode
29
=
*(1
0.00216s)(l
0.0268j)
plot, G(s)
is
written
(9-63)
496
Frequency-Domain Analysis
of Control
The Bode
Systems
is
is
is
about
50.
Chap. 9
of the system
domain parameters.
Table 9-3
9.11
Peak
Rise
Overshoot
Time
Delay
Time
BW
(henrys)
(%)
(sec)
(sec)
(radjsec)
<o p
Gain
Margin
Phase
Margin
(dB)
(deg)
0.1
41
0.4
0.5
48
1.78
40
4p
0.01
11.5
0.6
0.4
50
1.18
35
24
50
The constant
M and N
loci
function
is
is
M(s)
Q)
<M
f9-64)
and TV loci derived earlier and the Nichols charts of Figs. 9-27
and 9-28 cannot be applied directly. However, we can show that with a slight
the constant
still
P( S )
Comparing Eq.
G (s)H(s)
P(s)
(965)
we have
M(s)H(s)
(9-66)
Information on the gain margin and the phase margin of the system of Eq.
can be obtained in the usual fashion by constructing the Bode plot of
(9-64)
G(s)H(s). However, the G(jca)H(jco) curve and the Nichols chart together
P(jco) are related through Eq. (9-66), once the plots for P(co)
do not
and
versus co and
versus co are
and
<f>,(co)
Sec. 9.12
m (co)
9.12
IPjjco)
Frequency Domain
497
(9-68)
/H(jco)
Domain 4
in
that the sensitivity of a transfer function with respect to a given parameter can
We
be clearly interpreted.
shall
plot
chart can be utilized for analysis and design of a control system based on sensitivity considerations.
Consider that a control system with unity feedback has the transfer function
M^ = W)=TTW)
The
sensitivity
*"
{S)
_
~
dM{s)jM(s)
dM(s) G(s)
(9 ' 69)
is
defined as
,
q /U)
(y
dG(s)IG(s)
or
qm(\
(Q-iu
^ = r+W)
In general,
it is
a design standpoint,
we have
is
(9 " 72)
s.
From
form:
'
5*<'>l
(9 " 73)
iT+Wl^*
is
is
written
G(jco)\>l
(9 . 74)
The
is
at the
( 1, jO)
infinite.
498
Frequency-Domain Analysis
of Control
Systems
Chap. 9
/ImC
G(.s)-plane
Nyquist
plot.
Equation (9-74) and Fig. 9-31 also indicate clearly that for low
sensitivity,
the magnitude of G(jco) should be high, which reduces the stability margin.
among
control systems.
Although
in the frequency
domain,
in general, the
Nyquist plot
is
is
again
awkward to use
more convenient
for
for
Equation (9-72)
is
written
g-'(Jto)
TO)
(9-75)
G-^ja)
which clearly indicates that the magnitude and phase of Sffijco) can be obtained
by plotting G'^jco) in the Nichols chart and making use of the constant
loci
for constant sensitivity function. Since the vertical coordinate of the Nichols
chart
is
in decibels, the
the G(ja>)
G"
'
C/co)
is
dB
/g-'(ja>)
As an
= - G(jco)
= - /G(jco)
1
dB
(9-76)
(9-77)
Sec. 9.12
Frequency Domain
499
dB
Fig. 9-32.
Determination of the
sensitivity function
Sq
Figure 9-32 indicates several interesting points with regard to the sensitivity
The sensitivity function Sg approaches
dB
peak value of
dB
is
reached by
is
S%
most
at co
sensitive to the
frequency and more generally in this frequency range. This result is not difficult
to comprehend, since from the Nichols chart of Fig. 9-30 it is observed that
the stability and dynamic behavior of the closed-loop system
much
When
co p ,
more
directly
on the
is
is
which
is
raised in the
dB and
80, the
500
Frequency-Domain Analysis
of Control
G'
Chap. 9
Systems
{ja>) also
tivity is infinite.
In this section
may be
is
REFERENCES
1.
2.
1953.
3.
B. H. Willis and R.
of Regu-
lator Problems,"
262-267,
July 1965.
4.
PROBLEMS
9.1.
transfer function
is
shown
in
Fig. P9-l(a).
(a)
Compute
(b)
A zero
is
is
as
shown
P9-l(b);
in Fig.
how
../
.)">
s-plane
x-plane
45
s*
4?
-O-
-2
(b)
(a)
Figure P9-1.
Problems
Chap. 9
(c)
Another pole
is
on the negative
inserted
how
is
501
but at a
the band-
width affected ?
9.2.
The
specification given
on a
is
that the overshoot of the step response should not exceed 25 per cent.
(a)
What
resonance
(b)
9.3.
damping
ratio
and peak
p?
and
(o p
/ max
xT~~
o-As
-<><H
10
s(s
2)
X^_
i
(b)
~~x
A
\r*
,
10
+2s
s(s
2)
jy
(c)
+0.5x
10
s(s
(d)
Figure P9-3.
+ 2)
502
Frequency-Domain Analysis
(e)
Systems
of Control
Chap. 9
\M\
|-versus-co
curve
is
as
|1.5
0.8
^*~~~
a.
rad/sec
(e)
The closed-loop
(s)
(b)
R(s)
(1
is
given by
a> of
the system.
(c)
9.5.
The open-loop
G(i)
s(\
(a)
(b)
(c)
9.6.
is
it
K so
+ 0As)(l + s)
of the system
is
1.4.
Determine the value of so that the gain margin of the system is 20 dB.
Determine the value of K so that the pliase margin of the system is 60.
The open-loop
is
+ tS )
G(s)
uw = s(l + s)(l + 0.01s)
T so
an
infinite
gain margin.
9.7.
The open-loop
G(i)
=
s(l
is
+ 0.ls)(.l + 0.0015)
Determine the value of K if the steady-state error of the output position must
be less than or equal to 0.1 per cent for a ramp function input With this value of
K, what are the gain margin and the phase margin of the system? Plot G(s) in
the gain-phase plot and determine the resonance peak
p and the resonant
frequency
9.8.
a>.
A random compensation
in
Problem
9.7,
so that
network
is
added
reads
Problems
Chap. 9
G(s)
=
5(1
where
503
K(l + 0.01675)
+ 0.00222j)(1 + 0.1s)(l + 0.0015)
of G(s). Evaluate
p , 0i p , the gain
9.9.
co(rad/sec
0.1
Same
to cj
10 3
10 4
slope
-0
/G(/co)
Figure P9-9.
(a)
Find the gain margin and the phase margin of the system.
open-loop transfer function is changed to e~ Ts G(s), find the value of T
so that the phase margin of the system is 45. Then find the value of T so
that the gain margin is 20 dB.
What is the velocity error constant of the system in part (a) ? in part (b) ?
(b) If the
(c)
10
Introduction to Control
Systems Design
10.1
Introduction
Design of control systems represents an interesting and complex subject in control systems studies. In a simplified manner the design problem of control systems can be described with the aid of the block diagram of Fig. 10-1. The figure
u(0
Fig. 10-1.
c(f)
Controlled
process G
Control
vector
Controlled variable
vector (output vector)
variables,
is
vector
c(f)
simplified
Once
is determined,
usually needed to generate this control from the reference inputs
the state vector x(r) or output c{i). Figure 10-2 illustrates the block diagram
a controller
and
is
back.
is
Fig. 10-2
is
is
504
is
Introduction
Sec. 10.1
u(0
lit)
Controller
505
c(0
Controlled
process
x(f)
It is interesting to give
trol
state feedback.
a brief review on the history of development of conmay help gain perspective in the understanding
of the subject.
The early stage of the theoretical development of the design of control systems was characterized by the works of Nyquist, Hall, Nichols, 2 and Bode,
who developed such classical methods as the Nyquist plot, Bode diagram, and
Nichols chart. A unique feature of these methods is that they are all graphical
techniques which are conducted in the frequency domain. As was pointed out
earlier that in the design of control systems, it is the time response that is of
1
importance, rather than the frequency response. The use of the frequencydomain techniques is simply due to the fact that the graphical techniques are
convenient to apply.
The
classical design
first fixing
the configura-
the system
The proper
direct
of control systems is very much a trial-and-error proa distinct disadvantage of the method, since it does not indicate
whether a solution even actually exists for the design problem at the outset. It
The
classical design
position. This
is
506
r(f)
-6/\
Chap. 10
e(0
Controller
u(t)
c(r)
Controlled
process
(a)
r(0
e(f)
/\
u(f)
Controlled
process
c(0
Gp
Controller
Gc
(b)
Fig. 10-3.
of compensation,
is entirely possible that the design requirements are so stringent or may even
be
contradictory so that they cannot be satisfied by any system configuration or
controllers that are physically realizable. Even when a solution does exist, the
For
instance, gain
in 1950 made posof control systems to be carried out in the ,?-plane. The main
advantage of the root locus method is that information on frequency-domain as
well as time-domain characteristics can be derived directly from the pole-zero
sible the design
The work by Norbert Wiener 3 in the late 1940s opened a new horizon to
the design of control systems. Wiener introduced not only the statistical considerations of control systems but also the idea of the performance index. For
the first time, the design engineer was able to start from a set of design criteria
Introduction
Sec. 10.1
is
507
He
analytical procedure.
optimum
is
In
many
may be random
in nature.
and
Unlike
on the antenna
is
best
described by some probabilistic function rather than by a sine wave or any other
deterministic signals. The main difference between a deterministic signal and
a
random
signal
is
what
a given time.
The principle of Wiener's optimization technique is demonstrated by the
block diagram shown in Fig. 10-4. The design objective is to determine the
is
it
closed-loop transfer function C(s)/R(s) of the system such that the error between
the desired output and the actual output is minimized. In Wiener's statistical
mance
index, /; that
is,
r(0
7\
Noise
= lim
is
(10-1)
\t) dt
Process
Controller
c(0
Gp
'
/x
rs
(0
Signal
Desired system
(model)
Fig. 10-4.
zation technique.
for using the mean-square error as the performance index is that the
minimization of this particular performance index induces an analytical design
procedure which makes use of the mathematical functions already defined in the
theory of probability and statistics. However, in practice, a system that is
The reason
optimum
in the sense of
minimum mean-square
it is
error
not
may be
errors than
on smaller
ones.
difficult to see
from Eq.
emphasis on large
508/
Chap. 10
shown in Fig. 10-4 can be used for the analytirandom inputs as well as systems with deterministic
inputs.
mance
When
is
f"
(0<ft
f\<t)\dt
Jo
[t\e{t)\dt
Jo
In fact, the
is
the
from the
At approximately the same time the analytical design principle and techniques were being developed, Truxal 5 proposed a synthesis procedure through
pole-zero configuration in the j-plane. The synthesis still makes use of the conventional design specifications, such as the relative
bandwidth, and
rise time,
damping ratio
error constants,
Based on the
design specifications, the closed-loop transfer function of the control system is
first determined, and then the corresponding open-loop transfer function is
signals are deterministic.
found. The advantage of this synthesis method over the classical frequencydomain design is that the designer is able to determine if the given set of specifications are consistent at the beginning of the design, so the amount of guesswork
and
trial and error are cut to a minimum. Furthermore, Truxal' s synthesis starts
with the closed-loop transfer function and then works toward the transfer func-
tion of the controller, whereas the frequency-domain design starts out with the
controller and then works toward the closed-loop transfer function to see if the
design specifications are satisfied.
It is very difficult to pinpoint exactly when the modern control systems
theory was inaugurated. In fact, the mathematical foundation of certain aspects
of modern control theory can be traced far back to works that were completed
some seventy years ago. For instance, the state-variable approach to linear sys-
tems
is
well
known
thesis,
widely
known
were
much
time.
The launch of the space age has placed a challenge to the control engineer
new methods of design of more complex control systems and to meet
more rigid requirements. The control engineer soon discovered that the conventional design was no longer adequate and rigorous enough to handle the complicated problems encountered in modern fire control systems, autopilot systo find
Sec. 10.1
Introduction
509
terns, missile
many
Modern
others.
multiple inputs
trol theory.
but
is
truly
modern control design contains, first, the development and formulation of the
mathematical theory, and second, the application of the mathematical principles to practical design problems. Indeed, at present,
a gap
still
and
it is
exists
in a prescribed sense.
< <
is
f so that the state vector x(t) is brought from the initial state
to the desired final state x(^) in the shortest possible time, subject to the
interval
The problem
is
usually
/ can be
integral
/=
CF[x(t),u(t),t]dt
(10-2)
dt
set at unity, so
(10-3)
ta
and the
final times.
As another example,
if it is
(t
t ), while keeping the
f
magnitudes of the controls within reason, the following performance index may
be used
C
" [x(0
f
x,,
xJ'Q[x(/)
xj
u'(0Ru(0} dt
(10-4)
mic programming,
maximum or minimum
controllability,
and
observability.
510/
Chap. 10
In this section
is
general idea of the basic problems involved in the design of control systems be-
on
The
this
analytical
design method and certain aspects of optimal control are covered in Chapter
10.2
1 1
Systems
In this section the classical design of control systems will be carried out in the
frequency domain and the .-domain. The designs are mainly affected by means
of Bode plot, Nichols chart, and the root locus.
To
let
us consider the
following example. Let us begin by considering the transfer function of a controlled process
,(1
+ ,X
*+
0-0125.)
10 " 5 >
which is
unity.
state error,
we can
find the
minimum
value of
K in
in
Chapter 6 on steady-
order to
error
fulfill this
requirement
steady-state error
Therefore,
e ss
100.
lim
_ < 0.01
(10-6)
it is
easy to
show that the system is unstable for all values of A" greater than 81. This means
that some kind of compensation scheme or controller should be applied to the
system so that the steady-state error and the relative stability requirements can
be satisfied simultaneously. It is apparent that this controller must be able to
keep the zero-frequency gain of s times the open-loop transfer function of the
compensated system effectively at 100 while maintaining a prescribed degree of
relative stability.
The
domain
is
best
by the Nyquist plot of Gp (s) shown in Fig. 10-5. In practice, we seldom use the Nyquist plot but rather the Bode plot for design, because the latter
is easier to construct. When K = 100, the system is unstable, and the Nyquist
plot of G p (s) is shown to enclose the ( 1,7*0) point. Let us assume that we desire
to realize a resonance peak of
p = 1.25. This means that the Nyquist plot of
= 1 .25 from below. If
Gp(s) must be tangent to the constant-M circle for
is the only parameter that we can adjust to achieve the objective of
p = 1.25,
Fig. 10-5 shows that the desired value of K is 1. However, with this value of K,
illustrated
Sec. 10.2
Systems
511
i/ImC
G-plane
Constant
M locus, M =
.25
ReG
Fig. 10-5.
Gp{s) =
K/[s(l
s)(l
0.01255)].
not
is
only
sec" 1 ,
satisfied.
shows that in order to simultaneously satisfy the transient and the steady-state
requirements, the Nyquist plot of Gp (s) has to be reshaped so that the highfrequency portion of the locus follows the
1 plot and the low-frequency
portion follows the
100 locus. The significance of this locus reshaping is
K=
K=
M=
10-5,
it is
two
com-
pensated locus
1.
K=
from the
100 locus and reshaping the locus in the
region near the resonant frequency ca while keeping the low-frep
quency region of Gp(s) relatively unaltered.
Starting
2.
Starting
from the
portion of
Gp (s)
K=
co p relatively
K =
unchanged.
100
512
Chap. 10
G p (s)
is pushed in the
added to the system in the positive direction in the proper frequency range. This scheme is
basically referred to as phase-lead compensation, and controllers used for this
purpose are often of the high-pass-filter type. The second approach apparently
In the
first
is
K=
trajectory in the
is
with
K=
is
G p (s)
is
filters.
Figures 10-6 and 10-7 further illustrate the philosophy of dssign in the frequency domain using the Bode diagram. In this case the relative stability of the
system is more conveniently represented by the gain margin and the phase
aa
it is
80
60
'A
40
^>
20
au3
a.
-20
<
-40
-90
^
ft.
180
-270
<*
O
<o
rt
J=
.0*
100
10
0.01
co rad/sec
Fig. 10-6.
compensation.
K/[s(l
?)(!
0.0125.$)]
with phase-lead
1000
CQ
Systems
Sec. 10.2
513
80
60
3
40
<1
20
<D
T)
a
H
<
- 20
-40
-90
^-v
180
d.
IS
-270
o
5
-360
43
a.
1000
0.01
cj rad/sec
Fig. 10-7.
Bode
plot of
Gp (s) =
#/[,$( I
s)(\
0.0125.*)]
with phase-lag
compensation.
If,
instead of adding
more
positive phase to
the amplitude
on the phase
shift
is
affected at a sufficiently
due to the phase-lag compensation is negligible at the phaseThus the net effect of the compensating scheme is the
crossover frequency.
improvement on the
It will be shown
G
where for a high-pass
this transfer function,
filter,
it is
p^
now
>
{s)
form of
s
s
z,,
+ z,
+ Pi
(10-7)
filter,
<
Using
z,.
be
moved
K=
100
may
514
Chap. 10
i-plane
No compensation
K=0
K
-80
Fig. 10-8.
and
/, is
chosen to be
less
than z u but
for effective control, the two values are chosen to be very close to each other
and
are relatively small. Because of the closeness of z, and p u the portion of the root
Phase-Lead Compensation
Sec. 10.3
515
700
s-plane
No
compensation
With
'compensation
//
/(
K=
K=
K=
100
K=
*
-80
Fig. 10-9.
Gc(s)G(s)
Kl[s(\
5)(1
+ s)(l + 0.0125.$)]
0.0125*)] (Pi
<
and
).
10.3
lag characteristics
is
needed.
Phase-Lead Compensation
In Chapter 6 a simple phase-lead compensation scheme using the transfer function (1
Ts) was described under the name "derivative control." Although this
practical
l ,
516
Chap. 10
-WW
Sees
zero
Sccs
infinite
impedance
impeda
ice
series
com-
The
is
impedance
is infinite.
transfer function of
This assumption
is
is
zero,
2 {s)
E,{s)
Ri
R 2 + R,R 2 Cs
R R 2 Cs
+R
2
(10-8)
or
E2 (s)
E^s)
Ri
+R
R,Cs
(10-9)
RjR 2 r
Cs
+ r7Tr~
_i_
Let
/?,
+R
a>
R2
(10-10)
and
R^R 2 c
+R
R,
(10-11)
(s)
1
1
+ aTs
+ Ts
>
(10-12)
RC phase-lead
network
RC Phase-Lead Network
RC
shown
in Fig. 10-1
1.
By varying
located at any point on the negative real axis in the s-plane. Since a
may be
>
1,
the
always located to the right of the pole, and the distance between them
determined by the constant a.
zero
is
is
Phase-Lead Compensation
Sec. 10.3
617
/"
s-plane
Fig.
Pole
Zero
- i/r
- \/aT
E2 (j)/,(s)
=(l/a)(l
a7S)/(l
Ts).
The polar
E
E
2 (s)
x
{s)
is
+ aTs
+ T5
shown
(10-13)
values of a. For any particular value of a, the angle between the tangent line
and the
maximum
phase lead
<f>
as a approaches infinity.
The frequency
0\
co m
a.
<aj< "3
J Im
a
E 2 (fu)
E (/co)
plane
03
/^^^l/^
s\
CO
CO
\
<t>m\
mX
| co
co - \
CO -V ool
(s)\
(1
\
CO -*oo|
a2
a,
a[E 2 (.s)IE
aTs)l(l
a3
Ts).
Re
518
Chap. 10
Bode plot of the RC phase-lead network. In terms of the Bode plot, the
network of Fig. 10-10 has two corner frequencies: a positive corner frequency at co = 1/aT and a negative corner frequency at at == \\T. The Bode
diagram of aE2 (jco)/E (jco) is shown in Fig. 10-13.
RC
aE-,
20 dB/decade
20 log 10 a
slope
dB
J_
aT
aE-,
Fig. 10-13.
(1
Ts) (a
Analytically,
is
the geometric
>
=- (I
aTs)/
1).
0 and
co m
may
co m
yflogio
^j,
log 10
(10-14)
y)
Thus
(10-15)
To determine
the
maximum
phase lead,
<f>
we
aE 2 (jco)/
E^jco) as
Arg
-
2 (jcoT
EiUco)-
tan'
aTco
tan"
To
(10-16)
tan
1
Tco
(aTco)(Tco)
(10-17)
Phase-Lead Compensation
Sec. 10.3
When
519
q>
a),,
= -JL-
(10-18)
VaT
ia-mUa) = a_^
1
2v
or
sin
This
last
expression
is
= JTI
is
RC network of Fig.
as follows.
It is
specifications
1.
Gp (s)
for the
re-
uncom-
set
The phase margin and the gain margin of the original system are
read from the Bode plot, and the additional amount of phase lead
needed to provide the required degree of relative stability is determined. From the additional phase lead required, the desired
value of m is estimated accordingly, and a is calculated from Eq.
<j)
(10-20).
3.
4.
if
not, a
new value of
<f>
steps repeated.
phase-lead network
is
established
T.
Ic
,
.5
a s
OS
^i
3
'
5"
xf
u
CQ
e
o
o
0/
u
<u
5
O
o
M
520
Sec. 10.3
Example
by a
521
10-1
is,
~7
'
=K =
s
(10-21)
constant
(X
in the original system has also been eliminated for the present design
block representing the controller, in case it is needed, is inserted
between the two amplifiers. The parameters of the system are given as
The tachometer
problem.
RF =
10,000 Q.
K = 0.0125 volt/rad/sec
Km = 0.0125 newton-m/amp
b
Ra =
6.25
10" 6
kg-m 2
K =0.1 amp/rad
K = variable
s
B=
n =
The open-loop
800
a(s)
uncompensated system
is
written
K RF KKm jn
_
~~
RaJs 2 + Km K s
,. -.,
Kl "~^>
Substituting the numerical values of the system parameters, Eq. (10-22) gives
(s)
Ws)
The
_ 2500*
~
s(s + 25)
,
no
KW' Li)
2.
final steady-state
output
velocity. In other
be
less
The following
1
compensation
l^~
Since 9 r (s)
lim s*(s)
1/s 1 ,
lim s
s-0
j-,0
/*\
+ [Po(s)/(j)]
(10-24)
0.01/*
(10-25)
522
Chap. 10
1.5
Without compensation
',
"'2?^ 2s
+ 0.0106s
MO
With phase-lead compensation,
Gc (s) =
0.2
tn^tlir
0.4
0.3
Time (seconds)
Fig. 10-15. Step responses of the sun-seeker system in
Example
10-1.
44.4 per cent. Figure 10-15 shows the unit step response of the closed-loop
system with
2.
3.
K=
1. It is
is
quite oscillatory.
The Bode
plot of
system.
4.
Let us choose to use the phase-lead network of Fig. 10-10 as the controller.
Then, the transfer function for the controller of Fig. 10-14 is
= G (s)
(.s)
+aTs
Ts
As mentioned
lead network
assumed
(10-26)
to be absorbed
Phase-Lead Compensation
Sec. 10.3
40
I
0) _
II
S23
100
s(l+ 0.04s)
a(s)
20
dB
.,
G
cW-(i+ 0.0106s)
(1+ 0.0262s)
100(1 + 0.0262s)
20
s(l
+ 0.04s)(l + 0.0106s)
I
^^
-90
Phase of compensated system
180
_l
10
Itl
-'
60
50
'''I
100
L_
1000
co rad/sec
Fig. 10-16.
Example
Bode
plots of
in
10-1.
less.
is
no longer
if
25,
difficulties in
phase-lead compensation
may become
amount of phase
lead,
it is
ineffective.
difficulty,
essential to include
for the inevitable phase dropoff. Therefore, in the present design, instead of
selecting a
m of 20,
we let
sin0
from which we
= 0.422 = a+
(10-27)
1
get
2.46
(10-28)
<f>
524
Chap. 10
(a)
(f>
m occurs
at co'c
This
may be
calcu-
is
lated:
20 log 10 a
20 logi 2.46
7.82
dB
(10-29)
(b)
~ =
7 oo
floC/Q>)
From
-3.91 dB
found to be
co m
--=
60 rad/sec.
Now
we have
= VaOJm = \/2A6
-j
is
(10-30)
x 60
94 rad/sec
(10-31)
Then
^j,
The parameters of
38.2 rad/sec
now
17(f)
(s)
is
11+ 0.02625
+aT,9
(10-32)
Ts
2.46
(10-33)
+ 0.0106*
Since it is assumed that the amplifier gains are increased by a factor of 2.46, the
open-loop transfer function of the compensated sun-seeker system becomes
6150(5
flofr)
*{s)
s(s
38.2)
25){s
u "-34)
94)
In Fig. 10-17 the magnitude and phase of the original and the compensated systems are plotted on the Nichols chart. These plots are obtained by taking the data
from the Bode plot of Fig. 10-16. From the Nichols chart, the resonance peak,
of the uncompensated system is found to be 1 .88, or 5.5 dB. The value of
p with
directly
compensation
is
is
is
point
and the
rise
time
is
be
in Fig. 10-15.
Note
that
may
is
also reduced.
cent,
is
in certain systems
On
critical.
we
we
To demonstrate the
a to be 5.828. The resulting controller
select
-160
-140
-120
Phase (degrees)
Fig. 10-17. Plots of
GO)
Example
10-1.
52S
1.0
300
Example
oo
'/CO
K
5-plane
,
/sT=2500-
K=
/48.3
= o
a:
-25
K =2500-
/48.3
'
K
oo
(a) G(s)
s(s
Fig. 10-19.
(a)
526
+ 25)
in
Example
10-1.
Phase-Lead Compensation
Sec. 10.3
s-plane
A:
527
/to
250oY (-36.7+/61.7)
K=
AT
=2500
K=
94
14.7
AT
=2500
(-36.7-/61.7)
(b) G(s)
s(s
K{s + 38.2)
+ 25)(s + 94)
controller.
Gc
The
^ " 5.828
is
1
1
+ 0.03427*
+ 0.00588J
(10-35)
shorter
cent.
Using the magnitude-versus-phase plots and the Nichols chart of Fig. 10-17, the
closed-loop frequency responses for the sun-seeker system, before and after compensation, are plotted as
To show
shown
in Fig. 10-18.
and
after
It is clear
from
these diagrams that the phase-lead compensation has the effect of bending the complex
root loci toward the left, thus improving the stability of the system. The eigenvalues of
45.5, 36.7
the compensated closed-loop system are at s
j61.7, and 36.7
y"61.7. Therefore,
system
is
the
damping
ratio of the
In general,
on the results of the Bode plot design. In the following section, we shall illustrate a
procedure of designing phase-lead compensation using the root contour method.
528
Chap. 10
The root contour method can be used for the design of control systems.
Let us use the sun-seeker system studied in Example 10-1 to illustrate the design
procedure.
The open-loop
0M = G(s)
{} =
a(j)
The
uncompensated system
250
s(s
HO-36^
(1UJ&)
25)
+ 25s +
is
is
2500
(10-37)
^ --
2500(1
G(s)
s(s
25)(1
aTs)
Ts)
Q 38)
(1 "
such as the phase margin and gain margin, it is more convenient to specify the
relative positions of the complex eigenvalues.
To begin with the root contour design, we first set a to zero in Eq. (10-38).
25)(1
do not contain
T.
We
is
Ts)
2500
(10-39)
(10-39)
by the terms
have
1+
This equation
25
^+lL + 2 500 -
10 " 4 )
of the form of
s(s
25)(1
Ts)
2500(1
aTs)
(10-41)
Now we
This
is
not contain
a.
We
have
2500aTs
1+
Q
U
+ s(s + 25)(1
+ Ts) + 2500 ~
1Q 4Z)
(iU
Phase-Lead Compensation
Sec. 10.3
529
ico
i-plane
F=0*
-/48.4
/43.3(r = 0.0133)
T=cc
o<-r
r=oo
7"=oo
25
-/43.3(r= 0.0133)
^=0*-
Fig. 10-20.
0,
This equation
loci of
and
is
/48.4
+G
2 {s)
0,
10-1 with
T, the root
Eq. (10-41) can be obtained based upon the poles and zeros of
2500aTs
25X1
Ts)
G 2 (s)
is
(10-39),
s(s
lie
(10-43)
'
2500
10-20, for a given T. In other words, the root contours of Eq. (10-41) as
must start from the points on the trajectories of Fig. 10-20. These root contours
end at ^ = 0, oo, oo, which are the zeros of G 2 (s). The complete root contours
of the system with phase-lead compensation are
From
T should
now
we can
little
530
Chap. 10
oo
\\
oo
00
/'<0
s-plane
Nt
36.7
2.46\ \
+/61.7_S*_J^^v\
a =
,T=0.0106
/\--T = 0.0133
- 12.5+/48.4^5?
a =
J 7"= 0.05
'
T=0
ly\
/^</
a<\
(t=2.5
7 = 0=
<
-25
Root contour on
r==
real axis
/
'*.
Fig. ip-21.
troller,
improvement
is
/48.4
Root contours of the sun-seeker system with a phase-lead con= (1 + aTs)l(l + Ts), Example 10-1.
c (s)
Effects
From
tems as follows
1.
gain-crossover frequency
increased.
is
is
usually improved.
2.
is
fre-
Sec. 10.3
531
are improved.
3.
4.
are increased.
5.
sation
is
Bandwidth considerations:
If the original
system
is
unstable, the
of a
bandwidth
become
objectionable. Also,
if
come
may
be-
value of a
2.
is
justified,
new
gain crossover
is
added to a much
The
desired
may
2.
The open-loop transfer function has two or more poles that are close
to each other and are close to the gain-crossover frequency.
The open-loop transfer function has one or more pairs of complex
conjugate poles near the gain-crossover frequency.
will illustrate
is ineffective.
532
Example
10-2
Chap. 10
feedback be
G> (s) =
It is
,(1
.1*X1
(1(M4)
+ 0.25)
2.
ramp
function input
Phase margin
is
0.01 in magnitude.
> 40.
dB
1000
Phase curves of
phase-lead
compensated system
II
000
Fig. 10-22.
effects
0.l5)(l
0.2j)]
and the
Sec. 10.3
From
533
G p (s) when K
100
is
shown
is
is
in Fig. 10-22.
unstable for
all
values of
K greater than
15.
is
unstable.
more phase lead is being added, the gainpushed to a higher value. Therefore, for this case, in
which the uncompensated system is very unstable at the outset, it may be impossible to
realize a phase margin of 40 by the phase-lead network of Fig. 10-10.
In a similar fashion, we may use the root contour method to illustrate why the
phase-lead compensation is ineffective in this case of a highly unstable system. Figure
10-23 shows the root locus diagram of the uncompensated system with the process
described by the transfer function of Eq. (10-44). When A^ = 100, the two complex
eigenvalues of the closed-loop system are at s = 3.8 +/14.4 and 3.8 j 14.4.
It is
clearly
shown
crossover frequency
is
also being
s-plane
/15
K=
'/7.07
(K=
100
15)
/5
K=100
1
a:
= o
K=
K=
rf
-22.6 -20
10
-5
-/5
-/7.07(AT = 15)
\ K=
\ K
100
-/is
Fig. 10-23.
Root
loci
of s(s
5)(s
10)
50AT
0.
534
Chap. 10
+ aTs
+ Ts
a>
(10-45)
Gc (s)Gp (s)
5000(1
s(s
5)(s
+ aTs)
10)(1 +
(10-46)
Ts)
s-plane
r=o
Fig. 10-24.
of s(s
Root contours of
5)(s
10)(1
Ts)
s(s
5)(s
5000(1
10)(1
0.
aTs)
Ts)
5000
0,
and
Phase-Lag Compensation
Sec. 10.4
First,
tion
we
set
we vary rfrom
while
zero to
infinity.
The
535
characteristic equa-
becomes
s(s
5)(s
10)(1
Ts)
5000
T=
Next,
become
T=
s(s
is
5)(s
(10-47)
10)
(1(M8)
points,
oo.
the trajectories
l{s) ~
5000
on which a
in Eq. (10-46),
written
s(s
5)(.s
10)(1
Ts)
5000(1
aTs)
(10-49)
Thus, as
we have
(10-47)
+
^ s(s +
5000aTs
5)(s
10)(1
+ 5000
Ts)
/x_
r
U2W
(10-49)
by the
we have
a varying,
= Vo
(lO-SO)
(iv-W)
start (a
= 0) at the poles of
5000ar,r
s(s
5)(.s
10)(1
Ts)
uuol)
5000
We shall
is
far
show in the following that for the system considered in this example, it
more effective to compensate it by a phase-lag network, or by means of an auxil-
10.4
Phase-Lag Compensation
In contrast to using a high-pass
may
use a low-pass
filter
filter
we
RC
network that may be used for the low-pass or phase-lag compensation of control systems. If we assume that the input impedance of
the network is zero and that the output impedance which
^i
the network sees is infinite, the transfer function of the
+
1
network is
E^> ,
1 + R t Cs
nf
llU
- 1 + (*,
+ R 2 )Cs
,(*)
b2
Let
C^
aT
=RC
(10-53)
and
Fig. 10-25.
RC phase-lag network.
R*
Rp + RD 2
t
<
(10-54)
K
'
536
Chap. 10
aTs
Ts
<
(10-55)
RC Phase-Lag Network
Characteristics of the
RC phase-lag network.
l/T. As shown in
The
s ==
transfer function
is less
always located to the right of the zero, and the distance between them
mined by
is
is
deter-
a.
iu>
5-plane
aT
<
Ts),
(1
aTs)j(l
of a phase-lag network.
RC phase-lag
When we
= jco,
let s
E2 (jco) =
EiUm)
The polar
values of a,
(a
to the semicircle
of the network.
more
+ jooaT
>
<
a2
1),
>a
shown
co ml
>
oo ml
origin
>
(10-56)
(<f>
<f>
<f>
90
is
and the
As
<
l+ja>T
>
any value of a
as a approaches zero.
m occurs, co m
As
the
increases; that
is,
Phase-Lag Compensation
Sec. 10.4
*-
-/
537
Re
Im
Fig. 10-27. Polar plots of
Bode plot of
the
E2 (s)IE
RC phase-lag
is
shown
(s)
(1
aTs)/(l
Ts)(a
<
1).
network. The
in Fig. 10-28.
corner frequency at
<a
1/aT,
aT
90
^ ~ "^
*
'i
'
u **?
"~
90
Fig. 10-28.
Ts) (a
<
(1
aTs)l(\
538
readily be
satisfies
Chap. 10
lag
<f>
is
to increase
the phase in the vicinity of the gain-crossover frequency while keeping the
magnitude curve of the Bode plot relatively unchanged near that frequency.
However, usually in phase-lead design, the gain crossover frequency is increased
because of the phase-lead network, and the design is essentially the finding of a
compromise between the increase in bandwidth and the desired amount of
relative stability (phase margin or gain margin). In phase-lag compensation,
however, the objective is to move the gain-crossover frequency to a lower frequency while keeping the phase curve of the Bode plot relatively unchanged at
the gain-crossover frequency.
Bode
plot
is
outlined as follows:
1.
The Bode
pensated system
is
uncom-
is
set
The phase margin and the gain margin of the uncompensated system
are determined from the Bode plot. For a certain specified phase
margin, the frequency corresponding to this phase margin
on the Bode
plot.
The magnitude
is
found
system must be located at the point where the specified phase margin
is
3.
found.
To
gain-crossover frequency,
the
co'c ,
down
dB
to
at the
new
prescribed
at the
\G P Uo>'c)\
= -20 log,
dB
<
Gp (s);
then
(10-58)
from which
4.
is
io-io>.ow>i/2o
<
(10-59)
proper value of
T to
c,
is
not affected.
Phase-Lag Compensation
Sec. 10.4
This
is
539
tude curve will bring change to the phase curve, and vice versa.
With reference
shown in Fig.
quency, 1/aT,
cd'c ,
is
apparent that
it is
if
new
network
gain-crossover frequency,
co'c
the bandwidth of the system will be too low, causing the system to
that
is,
rad/sec
aT
10
10
(10-60)
Therefore,
5.
The Bode
rad/sec
(10-61)
is
investigated
and
T are
sub-
phase-lag network.
Example
In
example we
10-3
this
shall design
Example
Sift
flC(s)
The
s(s
dO-62)'
25)
1.
2.
The Bode
margin
is
1 is
shown
K>
in Fig. 10-29.
output
As
only 25.
Fig. 10-29
let
is
E^7)
From
K=
final steady-state
requirement of
it is
= G^S) = T +
<
gain-crossover frequency
540
Chap. 10
60
ii
i
i
s(l+0.Q4s)
a(s)
dB
Phase margin of
compensated system
|
180
0.4
0.1
1.0
Phase margin of
uncompensated system
compensated system
I
J-U
i_i_
i_i
2.0
100
co rad/sec
Fig. 10-29.
Example
Bode
plots of
in
10-3.
it is a safe measure to choose this new gain-crossover frequency somewhat less than 25 rad/sec, say, 20 rad/sec.
From the magnitude plot of {jco)l,{jco), the value of 6 (jco)/tx,(jco) at Co'c = 20
gain-crossover frequency,
rad/sec
of 14
CO'c
is
dB
14 dB. This means that the phase-lag controller must provide an attenuation
at this frequency, in order to bring the magnitude curve down to
dB at
(10-64)
dB
is
realized.
In order that the phase lag of the controller does not appreciably affect the phase
at the new gain-crossover frequency, we choose the corner frequency 1/aT to be at 1
20 rad/sec. Thus
decade below co'c
CO'c
5r=To
==
20
Io
,,
2rad/sec
(10-65)
which gives
-=
= 0.4 rad/sec
(10-66)
Phase-Lag Compensation
Sec. 10.4
The
U(s)
E
and the open-loop
(s)
a(s)
s(s
541
is
= + 0.5^
+ 2.5s
1
(10-67)
0q(s)
compensated system
is
5(M.s + 2)
+ 0.4)(j + 25)
-_.
nft
(iv-ob)
The Bode plot of the open-loop transfer function of the compensated system is
shown in Fig. 10-29. We see that the magnitude curve beyond CO = 0.4 rad/sec is attenuated by the phase-lag controller while the low-frequency portion
new
is
not
is
is at 25 rad/sec. The phase margin of the comdetermined from Fig. 10-29 to be about 50.
pensated system
is
the uncompensated
of the system
is
is
It is
The unit step responses of the uncompensated and the compensated systems are
shown in Fig. 10-31. The effects of the phase-lag controller are that the overshoot is
reduced from 44.4 per cent to 22 per cent, but the rise time is increased considerably.
latter effect is apparently due to the reduction of the bandwidth by the phase-lag
The
controller. Figure 10-31 also gives the step responses of the system
7* of
when
the value of
is
T=
CO'c .
little effect
origin,
the region where the eigenvalues should be located to achieve the desired per-
two root
dB
-160
-140
-120
Phase (degrees)
Fig. 10-30. Plots of G(s) in Nichols chart for the system in
542
Example
10-3.
Sec. 10.4
543
1.5
Without compensation
With phase-lag controller, GJs) =
'
+
,
I
0.2
I,
Ts
+ Ts
0.4
0.3
Time (seconds)
Fig. 10-31. Step responses of the sun-seeker system in
Example
10-3.
K is
=
1
loci
of Fig.
phase-lag network hardly altered the phase plot near the gain-crossover fre-
of the compensated
may he
5.
Another explana-
1*1
s, is
a factor of
From
uncompensated system
is
I*,
is
+25|
2500
written
(10-69)
544
Chap. 10
the value of
K at s
is
given by
,,
1
._
l_
lg,|lJ 1
+0.4|lj +25|
1
(10-70)
500|j,+2|
or
|J,||J,
1*1
since the distance
Si to
from
s t to
+25|
500
0.4
will
2. This argument also points to the fact that the exact location of the pole
and the zero of the phase-lag network is not significant as long as they are close
to the origin, and that the distance between the pole and the zero is a fixed
desired quantity. In the case of the last example, the ratio between 1 jaT and l/T
is 5.
Based on the discussions given above, we may outline a root locus design
procedure for the phase-lag design of control systems as follows
Since the design will be carried out in the s-plane, the specifications on the
dampand other quantities such as rise time, bandwidth, and maximum overshoot, which can be correlated with the location of
transient response or the relative stability should be given in terms of the
the eigenvalues.
/CO
.s-plane
? = 0.562 \
\\
= 0.707
a:
= 0.2
;18
/12.5
K = 0.125X
K = 0.0625
=
25
a:
= o
4e
12.5
(a)
system.
Root
loci of G(s)
250OK/[s(s
25)]
of the sun-seeker
Sec.
Phase-Lag Compensation
0.4
545
s-plane
r
f =
= 0.562\
^-|
0.707^
(b)
500(5
2)/[s(s
25)(s
+ 0.4)]
2.
sated system.
loci
the value of
3.
Compare
the value of
is
ratio
lag controller.
4.
We may choose
T is
not
critical as
long as
many
it is
relatively large.
orders of magnitudes
Let us repeat the design of the system in Example 10-3 by means of the
root locus method just outlined. Instead of using the phase-margin specification,
we require that the damping ratio of the dominant eigenvalues of the closed-
loop system be approximately 56.2 per cent. This value of the damping ratio is
chosen so that we can compare the result of the root locus design with that
which was obtained independently by the Bode diagram method. The root locus
546
greater.
From
may be
it is
a
desired relative stability
K of the
of K is
attained by setting
The
Chap. 10
02
T=T
1
attained by setting
is
ratio of
K to
(s)
a(s)
The open-loop
_
~
500
s(s
Q) _
2500^(1
s{s
a(s)
{W ' U)
25)
compensated system
is
given by
aTs)
25)(I
Ts)
(10-73)
2500a^(j+ 1/aD
s(s
If the values of
+ 25)(s +
1/T)
(Us)^2500aK_
=
a(j)
j(j
25)
is
approximately
<
from the transient standpoint. Since K is necessarily equal to unity, to have the
right sides of Eqs. (10-72) and (10-74) equal to each other, a = 1/5, as is already
determined in Eq. (10-71). Theoretically, the value of T can be arbitrarily large.
However, in order that the bandwidth of the closed-loop system is not too small,
we must not make Ttoo large. By setting T
2.5, the root loci of the compen-
As an alternative, the root locus design can also be carried out by means
of the root contour method. The root contour method was applied to the sunseeker system earlier in the phase-lead design. The design carried out by Eqs.
(10-38) through (10-43)
and
<
Example
10-4
Consider the system given in Example 10-2 for which the phase-lead
compensation is ineffective. The open-loop transfer function of the
original system and the performance specifications are repeated as
follows
The performance
specifications are
1.
K =
2.
Phase margin
100 sec"
1
.
> 40.
Sec. 10.4
i-plane
;Q.2
Fig. 10-33.
K=
T=2.S
T=5
26.45
Root contours of
1.
3.
r=10
T= 10
T=5
T= 2.5
25.76
s(s
25)(1
Ts)
2500A:(1
(0.28+/4.4)
(-0.51)
(-1.04)
(-2.2)
aTs)
0;
1.
2.
r=2.5
T = 5.0
is
as follows
The Bode plot of Gp(s) is made as shown in Fig. 10-34 for K = 100.
The phase margin at the gain-crossover frequency, CO c = 17 rad/sec, is
approximately 45, and the closed-loop system is unstable.
The desired phase margin is 40; and from Fig. 10-34 this can be realized
if the gain-crossover frequency is moved to approximately 4 rad/sec. This
means that the phase-lag controller must reduce the magnitude of G p (jco)
to
dB, while it does not affect the phase curve at this new gain crossover
frequency,
G>'c
is still introduced by
a safe measure to choose the new gain-
co'c , it is
crossover frequency
alternative,
4.
lQ-\oMo>c')\no
10 -i.5
= o.032
(10-76)
This equation implies that the two corners of the phase-lag controller
548
Chap. 10
80
10Q
s{\
0.1 s)(l
+ 0.2s)
cj c
-180-
Phase margin of
compensated system
7 rad/sec
Phase margin of
uncompensated system
270'
J_
0.01 0.0112
0.35
0.1
1.0
2.5
100
cjI
w rad/sec
Fig. 10-34.
Example
Bode
plots of
in
10-4.
must be placed
1.5
dB
of attenuation.
5.
Using the guideline that the upper corner frequency of the controller,
1/aT, is placed at the frequency of 1 decade below the new gain-crossover
frequency, we have
1
3.5
(o'c
,,
^ = To=To =a35rad/sec
which gives
6.
The Bode
T=
Gc (s) =
by
1
sated system
7.
89.3
is
(10-77)
is
+ aTs
+Ts
It is
1
1
+
+
2.86j
(10-78)
89.3s
approximately 40.
= G (s)Gp (s)
c
+ 2.865)
+ 0.2^(1 +
100(1
s(l
0.ls)(l
89.3*)
(10-79)
systems are plotted on the Nichols chart, as shown in Fig. 10-35, These curves
show
260
220
240
180
- 160
140
100
120
Phase (degrees)
Fig. 10-35. Plots of G(i) of the
tems
in the
sys-
549
550
Chap. 10
resonant peak
dB
(1.41)
= 40 deg
= 10 dB
= 6 rad/sec
phase margin
gain margin
bandwidth
K=
When
100, the open-loop transfer function of the uncompensated system and
the compensated system may be written
G{s)
s(s
G0)
5000
5)0 + 10)
160Q
s(s
5)0
(10-80)
+ 0.35)
+ 0.0112)
10)0
s-plane
/7.07(A = 15)
/6.73 (A = 420)
(AT
= 3.2)
- 1.53+/3.335
(K= 100)-
1.44
K=0
+/2
K=0
10
50K
G(s)
s(s
1.6A"(s
G(s)s(s
Fig. 10-36.
Example
Root
10-5.
+ 5)(s+
loci
+ 5)(s+ 10)
+ 0.35)
10)(s
+ 0.0112)
(10-81)
Phase-Lag Compensation
Sec. 10.4
respectively.
5000
is
551
Eq. (10-80)
is
is
G(s) in
is
the pole and zero of the phase-lag controller are very close to the origin, as compared
5 and 10, the controller effectively reduced the loop gain of
with the poles at s
the system by a factor of 0.032. Figure 10-36 shows the root loci of the uncompensated
and the compensated systems. Again, the loci of the complex roots are very close to
Time (seconds)
Fig. 10-37. Unit step response of the system with phase-lag
in
Effects
Example
compensation
10-4.
From the
be summarized as follows:
1
2.
3.
is
decreased.
transfer
552/
function
Chap. 10
is
improvement
of the system.
4.
The
rise
10.5
usually decreased.
is
rise
time
is
is
usually
For additional
further increased.
Lag-Lead Compensation
We
usually improves the rise time and the overshoot but increases the bandwidth
On
overshoot and relative stability but often results in longer rise time because of
reduced bandwidth. Therefore, we can say that each of these two types of control has its advantages and disadvantages. However, there are many systems
that cannot be satisfactorily improved by the use of either scheme. Therefore,
it is
natural to consider the use of a combination of the lead and the lag con-
time,
may
combined
structure.
The
may
be
written
--(t^)(ttt#)
I
where a
tion if
>
and b
we assume
<
1,
1
1
is
<- lag
/a
<
1M2 >
is
compen-
Usually it is not necessary to cascade the lead and the lag networks of Figs.
10-10 and 10-25 for the realization of Eq. (10-82) if a and b need not be specified
independently.
ber of elements,
is
shown
in Fig. 10-38.
The
-vwv-
R2
C2
network
is
Lag-Lead Compensation
Sec. 10.5
2 (s)
G A)
M _~ e,(j)
~
Comparing Eq.
+ *,c,jX1 + R c s)
+ * ,C + tf C > + R^C^s
(i
1
(J?,C,
.
(10
83)
v
= ^C,
AT = * C
T T =R R CC
(10-84)
7\
From
(10-86)
(10-87)
RiRiCiCt
ab
(10-88)
specified independently.
10-5
The performance
< 10
- 89
>
specifications are
K =
100 sec"
v
Phase margin
1.
we have
abT T 2
2.
(10-85)
Thus
Example
553
we have
1
.
> 40.
Example
is reduced.
Let the series controller be represented by the transfer function
r<\ =
GM
For the
first
part
we
(1
(1
+ aT,s)(l + bT s)
+ ri ,xl + TlS)
In general, there
is
no
no9m
10 - 9 )
is
realized
by the network of
(10-88).
arrangement.
Let us
first
proper values of
frequency of
desired
new
alone
used.
is
Gp (ja>)
K=
100.
We
from 17 rad/sec
arbitrarily
gain-crossover frequency
is
3.5 rad/sec
when
Example
10-4, the
phase-lag compensation
margin of the
system should be improved to approximately 10. Using the phase-lag design technique,
we notice that the attenuation needed to bring the magnitude of G{jco) down to dB
at co'c = 6 rad/sec is 22 dB. Thus using Eq. (10-59), we have
b
'
io- 22 20
10- 1
'
=0.08
(10-91)
554
Chap. 10
80
60
gpX(S )'
40
s(l+0.\s)(l+0.2s
20
dB
-20
s(\
- 40 -
-60
Phase margin of
compensated system
90
h-
Phase margin of
uncompensated system
180
-270 J_
0.048
">,.
0.1
0.6
1.0
3.11
100
38.9
tj rad/sec
Fig. 10-39.
Bode
plots of
Placing \jbT2 at
j-jt
-j-jj
co'c
6 rad/sec,
we have
0.6 rad/sec
(10-92)
Thus
20.83
(10-93)
0.048
(10-94)
and
T2
The phase-lag portion of the
controller
1
Now we
is
described by
+ 1.667j
+ 20.83.?
is
(10-20), the
aa+ = 0.8518
phase
(10-95)
or
equal to the
sin0
(f>
is
maximum
58.3 deg
controller
is
Lag-Lead Compensation
Sec. 10.5
20 Iog,
20 log,
12.5
21.9
dB
555
(10-96)
is
found to be
at co
r,
a>m
(10-97)
38.:
and
aT
The
3.11
y
(1
c (s)
(1
where as a usual
+0.32s)(l
determined as
is
+ 1.6675)
+ 20.83*)
(10-98)
0.0257j)(1
been
dropped.
Figure 10-39 shows that the phase margin of the compensated system is approximately 40. The open-loop transfer function of the compensated system with K = 100
is
G(s)
The
= G (s)Gp (s)
c
4981(*
s(s
5)(s
3.11)(j
10)(.s
compensated system
is
+ 0-6)
+
38.88) (s
shown
(10-99)
0.048)
in Fig. 10-40. It
is
1.5
^._-"
1.0
System with
lag-lead
controller
0.5
I
1
Time (seconds)
Fig. 10-40.
controller.
apparent
566
Chap. 10
5-plane
AT
= 4981
-
/K=
4981
K=
42.627
38.88
Fig. 10-41.
the system in
lag-
lead controller.
that the step response of the system with the lag-lead controller is much improved
over that of the phase-lag compensated system. Not only the overshoot is smaller, but
is also greatly reduced. We may attribute these improvements
to the
addition of the phase-lead portion of the controller. It can be easily verified that the
bandwidth of the closed-loop system is now approximately 18 rad/sec, which is 3
5487
(*
+ 42.627X* +
4.159 +/10.556)(j
+ 4.159
-7T0556)
effectively canceled
10 - 10)
by the zeros
at
It
on the
we have been
first trial.
may
Sec. 10.6
10.6
Many
plex poles
is
that the gradient of the phase with respect to frequency near the
gain-crossover frequency
is
when two or
more simple poles of the transfer function are placed close to each other. With
reference to the system considered in Examples 10-2 and 10-4, the reason the
phase-lead controller
tem is
is
ineffective in
stability
of the sys-
because the phase of the process has a steep negative slope near the gain-
crossover frequency (see Fig. 10-22). When a process has complex poles, especially
if the poles are close to the imaginary axis, the design problem may become
more acute. We may suggest the use of a controller that has a transfer function
with zeros so selected as to cancel the undesired complex poles of the original process, and the poles of the controller are placed at the desired locations
in the s-plane. For instance, if the transfer function of a process is
G
in
= ^ +* +
io)
1(M01)
problems in the
have the transfer func-
stability
may
tion
Gc{s)
= /'+/+]
(1(W 02)
to the performance
10-42(b)
referred to as type 2.
is
With
2 (s)
+ 2RC s + R C,C s
+ (C, + 2C )s + R C C s
1
Ei(s)
When
these
E^s)
1
1
(10-103)
v
is
2R.Cs + C2 R,R 2 s 2
C(R 2 + 2R,)s + C 2R,R 2 s 2
it is
(10-104)
'
v
568
Chap. 10
H(-
-\MAr-
^c,
(a)
R2
-WAr-
-K-
(b)
Fig. 10-42.
Two
Type
network, (b)
Type 2 network.
type
T^S ^ R C^C
+ RCi
Cj+2C s
RC^ + RtCtCz
%
Ms)
'
If
'
both the numerator and the denominator polynomials of Eq. (10-105) are
form
2Ccos
we
(10-105)
col
(10-106)
=+
R*JC C2
(10-107)
(10-108)
Sec. 10.6
559
r
*"
= co
RjC.C,
-C +
2C2
'
2yc,C 2
(10-109)
2C 2 /C,
2VQ/C,
2tf
(10-110)
2,
The loci of the poles and zeros of Ez (s)IE {s) of Eq. (10-105) when C
u Ci,
and R vary individually are sketched in Fig. 10-43. When R varies, the numerator
and the denominator of Eq. (10-105) contain R in the form of R\ and the root
locus method cannot be applied directly to this nonlinear problem.
Fortunately,
l
s-plane
s-plane
0-C,
*-
/CO
c, =
C, -> oo
X >
>
<
/?C2
/CO
s-plane
2/?C2
/CO
s-plane
ii
'
C2 ^
C2
"*
c2
C2
'
C2
10
RC
-*o
RCy
=
**-<>
__J_
RCy
II
o
s
-_
/^Z^
,.
i_A
C2
/?-oo
/CO
s-plane
>C
(C*<
1)
0*-/?\
/?C,
/?-*oo
RC
RC,
C2 <
C,
(C
>
1)
network.
560
Chap. 10
equations that are of the form of Eq. (10-106) are of the second
easily. Therefore, the pole and zero loci of Fig. 10-43
show that the two zeros of the bridged-T network type can be either real or
in this case, the
The and
C The
.
COnz
(10-111)
C*/R R 2
1
(10-112)
(o
(10-113)
(o 2
2*.
2
The root
network
if
Example
10-6
loci
shown
in Fig. 10-43
(10-114)
v /?i/?2
can
still
in Fig. 10-44
is
selected to demonstrate
3>\ *M
R(s)
K(\ +
Controller
Fig. 10-44.
Example
10-6.
ATO+JOs)
(10-115)
0.2* -f 0.25.S 2 )
shown
transfer function
.5(1
The root
for
C(s)
lO.s)
+ 0.2s + 0.25s 2
5(1
is
shown
in Fig.
is
is
constant of
oscillate
1 1
sec.
about a
Thus
this small
level that is
Sec. 10.6
K
0.354 +/6.62
561
Let us select the type 1 bridged-T network as series compensation to improve the relative stability of the system. The
/CJ
t
K=
work should be
i-plane
j' 2
GAs)
S
+ 0.8s + 4
+ CO
(10-116)
2t P co op s
(10-117)
/?V C\Cz
-0.4+ /1. 96
(10-118)
K
One root
-
0.093
From
W
at
(K=
co,
(10-119)
2.7
(10-120)
and
1)
K = o:
c,=
-0.4-/1.96
The
;3
/6.62
^ BC
bridged-T network
+ 0.8s + 4
+ 0.384)(j + 10.42)
^ > XU
The open-loop
is
s1
(J
0.354
+ 2E =
Gc (s) =
-/5
(10-121)
compensated system
is
[AC(1
10s)]/[.s(l
0.25^)].
sated system
K=
Although the preceding design is carried out on the basis that the undesirable poles of the controlled process are exactly canceled by the zeros of the
controller, in practice this ideal situation is difficult to achieve.
One of the difficulties
lies in the fact that we cannot design the controller so
that its poles and zeros
will fall exactly at the specified locations. Another problem
is lack of knowledge
as to exactly where the poles of the process are. Usually, in the determination
of
the transfer function of a process, assumptions and approximations
have to be
made, so that the mathematical description of the system is never precise. In
562
Chap. 10
/cj
s-plane
K=\
S.33+/3.8
K=
--o
10.4
One
-
K=
Fig. 10-46.
Root
loci
root at
0.093 when
K=
5.33-/3.8
0-6 with
bridged-T controller.
is
realistic situation in
Let us consider that the design of the bridged-T compensation of the system
10-6 resulted in an inexact pole-zero cancellation, so that the transfer function of the controller is
in
Example
c (s)
+ 0.76s + 3.8
+ 0.366)0 + 10-4)
s1
=
(s
(10-123)
Since the complex poles of the controlled process are not canceled by the zeros
G c (s), the open-loop transfer function of the system becomes
of
w=
40K(s
G(s )
For
all practical
s(s
0J6s
3.8)
Q.1Xj
0.8j
10.4)0 2
4)
0.366)0
(10-124)
K=
1,
the closed-
Sec. 10.6
563
Uncompensated system
c(t)
10
Seconds
Fig. 10-47. Unit step responses of the
sated systems in
C()
R(s)
Example
40(,y
(s
0.093)O
10-6.
0-l)Cr
0.81 ^
0-76a
3.805)O
+ 3.8)
+ 10.66.S +
(10-125)
42.88)
Since the zeros of the open-loop transfer function are retained as zeros of the
closed-loop transfer function, Eq. (10-125) verifies that near cancellation of
poles and zeros in the open-loop transfer function will result in the
tion for the closed-loop transfer function. Therefore,
is
Alternatively,
in
which the
relative positions
564
Chap. 10
/CO
X?fC=o
s-plane
K=
(a)
'
/'co
>
5-plane
K=
K=
K=Q
(b)
P.K =
Fig. 10-48.
Root
cancellations.
REFERENCES
1
2.
Problems
Chap. 10
3.
565
Series,
4.
W.
J.
Synthesis,
McGraw-Hill
1955.
PROBLEMS
10.1.
The open-loop
s(l
The feedback
is
unity.
+ 0.2i)(l + 0.5s)
is
to follow a reference
ramp
(a)
given by
jr
G(s)
is
(b)
M, and bandwidth.
(1
+ 0As){\ + 0.08*)
is
inserted in the forward path of the system. Evaluate the values of the gain
Comment on
mance.
Sketch the root
(c)
10.2.
p,
loci
shown
is
in Fig. P10-2.
the
Acceleration constant
(b)
K = 5 sec -2
p <
1.5.
*-C(s)
Figure P10-2.
Design a
Sketch the
of the damping ratio (of the complex roots) and the bandwidth of the compensated system ?
10.3.
Figure P10-3
illustrates the
The system may be used for the positioning of a shaft by a command from a
remote location. The parameters of the system are given as follows:
= 0.01
/ = 0.05
ze
sec
oz-in-sec 2
566
Chap. 10
Motor
Motor and
electric
circuit
Ka
load
Km
Gc (s)
+Te S
Figure P10-3.
= 0.5 oz-in-sec
Km = 10 oz-in/volt
B
TL =
(a)
less
(b)
disturbance torque
Determine the minimum value of the error sensor and amplifier gain Ka so
that the steady-state error of 9 due to a unit step torque disturbance is
than or equal to 0.01
(1
per cent).
c {s)
+aTs
Ts
a>
Ka
is set
10.4.
response O (O when the input r (t) is a unit step function. (TL = 0.)
Repeat the design problem in Problem 10-3 with a phase-lag controller
1
c (s)
10.5.
Human
+aTs
+Ts
<
beings breathe in order to provide the means for gas exchange for the
respiratory control system is needed to ensure that the body's
entire body.
needs for this gas exchange are adequately met. The criterion of control is
adequate ventilation, which ensures satisfactory levels of both oxygen and
cavity
signals
is
The
objective
is
human
(b)
Assume
is
maintained
in the
Kf
is 0.
Chap. 10
Problems
Lungs
567
Circulatory system
Concentration
CO)
0.1
Desired
ventilation
V
Chemoreceptor
Figure PI 0-5.
10.6.
JR
where
(a)
187?*
200 ft-lb-sec 2
R is
Assume
(b) Let
the
strictly,
process
10.7.
is
the
Speed
governor
Power system
valve
generator
Reference
input
voltage
Tachometer
Figure PI0-7.
568
Chap. 10
output voltage
(a)
is
KT and indicate the location of the roots with the desired damping.
Determine the desired reference input voltage, with the value of KT set at
the value determined in (a), so that the generator speed is 1200 rpm
of
(b)
(TL
(c)
= 0).
mined
in (a).
(d) It is desired to
0.1 per
cent.
variable parameter.
10.8.
The phase-lock loop technique is a popular method for dc motor speed control.
A basic phase-lock loop motor speed control system is shown in Fig. P10-8.
Reference
speed
Motor
Eo
ym
control
circuit
*2
<0
Encoder
Counter
speed
Motor
MM
C
1(-
vwv-
t>
Em
Filter
Figure P10-8.
train
proportional to the phase difference between the reference speed and the
actual motor speed. This error voltage upon filtering is used as the control
signal for the motor.
follows:
Chap. 10
Problems
Kp = 0.0609 volt/rad
Ka =
K"
=
(K -tor
{Km
- I0 T
Vm (s)
s(\ + Tm s)
K = 5.73 pulses/rad
Ea_ __ R Cs +
E
RiCs
*i = 1.745 x 10 Q, C, = fiF
Phase detector
Motor
Motor
569
QM
transfer function
<
Encoder gain
">
0.04)
in
1_
Counter
system bandwidth.
Sketch the root locus diagram for the characteristic
equation roots as R 2
varies.
10.9.
a system
with tachometer feedback. Choose the tachometer gain
constant K, so that the
relative damping ratio of the system is 50 per cent.
How does the tachometer
feedback affect the bandwidth of the system ?
X^
R(s)
a*)
10
s(s+ I)
K,s
Tachometer
F
10.10.
gure P10-9.
R(s)
E(s)
KX^
-T
1
+ Ts
Figure P10-10.
10
s(l +0.1.S) 2
C(s)
570
10.11.
Chap. 10
shown
and
BL
K =
BL =
e
where
K =
b
0.023 oz-in-sec 2
36 oz-in/volt
Ra
6.92 oz-in/rad/sec
KL =
10 oz-in-sec
JL = 7.24 oz-in-sec
K K IR + Bm Ka = torque
2
in oz-in/amp,
(a)
Write the
Q>l,
m and
,
2857.6 oz-in/rad
constant
friction constant.
system shown
in Fig.
P10-ll(b) with 9 L ,
Derive the
transfer functions
Tm'U m e
,
+ o
WW
(b)
Amplifier
Controller
Motor
rr,
9m
Speed
transducer
(c)
Figure P10-11.
i>
'
*o,
Problems
Chap. 10
a>m(s)
(b)
The
571
co L(s)
and
is
Ea {s)
rately.
fier
speed error
when
the input e r
is
a step function;
(2) the
dominant roots of
damping
ratio of approximately 0.707 and (3) what should be the value of the input e r if the steadystate speed is to be 100 rpm? Sketch the root loci of the designed system
with
as the variable parameter.
10.12.
in
Problem
10-1
system parameters
Ra
Jm
36 oz-in/volt
_ K K + Bm =
Ra
b
KL = 28,576
JL
(a)
oz-in-sec 2
BL =
Kf =
oz-in/rad
7.24 oz-in-sec
Show that
6.92 oz-in/rad/sec
= 0.023
0.01
is
proposed
P10-12 shows the block diagram of the overall system. The phase-lag
is for the purpose of ensuring
zero steady-state speed error when a step input is applied. Determine the
Phase-lag
Amplifier
network
>
+ 0.953
5
-j-VVW-n
*-
Bridged-T
network
Speed
transducer
Figure P10-12.
11
Introduction to Optimal
Control
11.1
Introduction
From the discussions given in Chapter 10 we realize that the conventional design
of feedback control systems has many limitations. The most important disadvantage of the conventional design approach
trial
and
error.
The optimal control design is aimed at obtaining a system that is the best
possible with respect to a certain performance index or design criterion. The
state-variable formulation is often used for the representation of the dynamic
process. Since the optimal control problem is more of a mathematical development, the problem is stated as
Find the control u(t) that minimizes the performance index,
J =['' F[x{t),n{t),t]dt
J
(11-1)
to
x and
is
described by
i(0
f[x(r), u(r),
t]
(11-2)
straints
572
on the
state variables
Introduction
Sec. 11.1
573
equation of Eq. (11-2) describes the controlled process only, the statement of the
is accomdepends on
it
is
naturally
formed.
To
Minimum-Time Problem
minimum
its states
may be
is
the
condition
stated so as to minimize
f
>
(1 1-1),
initial
Thus, in Eq.
from some
dt
(11-3)
to
t]=\.
F[x(t), u(r),
states at the
equilibrium state, and the system should be able to return to the equilibrium
from any
problem is
state
initial state.
J
where
Q is a
=$
[" x'Qxdt
(11-4)
[" [x'Qx
*
where a constraint
is
u'Ru] dt
(1 1-5)
to
f"
I
<<,
(x
x)'Q(x
-x )dt
d
(11 -6)
574
Chap. 11
minimum-fuel
11.2
Analytical Design
and many
others.
One of the
is
criterion,
first
known
The design
as
minimum
integral-square-error design.
significant in that
it
complete design in a mathematical and analytical way. The only problems the
designer has to be concerned with are: (1) if the design has a solution, and (2) if
the solution can be physically implemented.
to the conventional
domain
or,
approach
The
is
more appropriately
is
is linear,
11-1.
is
The assumptions
we
~\
KO
eit)
Fig. 11-1.
The
uit)
Linear
controller
Linear
process G(s)
*-
cit)
The control
signal
1-1 is
is u{i),
portrayed in such a way that its specific input-output configuration is not yet
defined. This allows the flexibility of selecting the final controller configuration,
such as a series controller or a feedback controller, once the design is completed.
is
integrals:
= C e\t)dt
(11-7)
Ju
Cu\t)dt
(11-8)
where
e(t) is
referred to as the
(ISC).
Analytical Design /
Sec. 11.2
The
575
restriction
on such
criteria as
minimize Je
and
Ju
the problem
is
equivalent to
=J +
minimize J
where k
On
is
K= constant
k 2 Ju
(11-9)
we wish
if
to
minimize Ju
and
Je
then the problem
is
=K
constant
equivalent to minimizing
Ju
k 2Jt
(11-10)
where E(s)
is
e 2 (t)dt
= ~. f
E(s)E(-s)ds
(H-H)
e{t).
Let us formulate the problem as being that of finding the optimal control
for the system of Fig. 11-1 so that
=J +
e
k 2 Ju
= P [e
(t)
k 2 u\t)] dt
(H-12)
is
minimized.
= ^p
'
\E{s)E{-s)
k 2 U(s) U(-s)] ds
(1 1-13)
576
Chap. 11
With reference
to Fig. 11-1,
E(s)
= R(s) -
G(s)U(s)
we
[R(s)
(1 1-14)
get
G(s)U(s)][R(-s)
G(-s)U(-s)]ds
(11-15)
+ 2nj)_
U(s)U(-s) ds
J
U(s)=UJs)
where A
the
left
is
The
is
We
(11-16)
{s)
last condition is
= J + k Ju = /, + X(J + /,) + A V
(11-17)
where
Jl
Jl
J%
/4
= _L
db |
(**
{-
+ G GVBk Uok
ds
(11-18)
~ G* + GGVtWi ds
t1
M9
klUok
-GR + GGUokWi ds
(1
1-20)
(k*
simplicity,
k * U*V*
(<kl ^
5T I
2^7 /
For
~ R k ~ RGU k +
GCftUiUi ds
we have used G
>
(11-21)
for G(s),
Uok
for
for
Uok {s),
and
so on.
/.
is
In Eq. (11-17), J t is the optimum performance index, that is, the minimum
J3 since the integrand of J2
Equations (11-19) and (11-20) indicate that J2
identical to that of J3 with s replaced by s. Also, J4 is always positive, since
its
integrand
The
is
(1 1-17) to
be a
minimum
are
dJ
(11-22)
>o
(11-23)
A=0
and
d 2J
dk 2
-1=0
J2
and Eq.
+J = 2J =
}
(11-24)
(11-23) gives
J*
>o
(11-25)
Analytical Design /
Sec. 11.2
which
always
is
minimum J are
satisfied.
(1
Since
we have assumed
of Eq.
(1
j-plane,
1-26)
is
and the
C7,
that
that
must
all
and
(1 1-20),
ok
the poles of
all
all lie
1-24)
-GR + GGU
(k 2 Uok
577
)U ds
1
(11-26)
One solution
left
half of the s-plane. Since the contour does not enclose any singularity of the
integrand, the contour integration
zero
if all
is
zero.
Thus J3 of Eq.
(11-26)
identically
the poles of
X(s)
X(s)
The function
j-plane.
is
{k 2
GG)
is
(1 1-27)
is
written
+ GG)Uok - GR
(k 2
(1
1-28)
We let
Y(s)
Y(-s)
= YY=k 2 + GG
(1
1-29)
where Y is a function that has only poles and zeros in the left half of the j-plane,
and Fis Y with s replaced by s and has only right-half-plane poles and zeros.
is
defined:
Y={k 2 +
Y={k 2 +
The process of finding Y and
Y is
GG} +
(11-30)
GG}~
(11-31)
is
Equation (11-28)
is
now
written
X=YYU -GR
ok
We must now separate the right side of Eq. (1 1-32) into two parts,
is
one with
(11-32)
left
4 = YU
ak
-^
(11-33)
GR
CGR~
rGRT
_Y + + _ Y _
(11-34)
(11-35)
(11-36)
where
GR
_Y
GK
Y
578
Chap. 11
GR
X_
GR
YUok -
(11-37)
In this equation, the left-side terms have only poles in the right-half plane, and
the right-side terms have only poles in the left-half plane. Therefore, the left-side
Uak =T7 GR
Note that
(11-38)
optimum value
Uok
for
which
= j_
k 2 The
to determine
Table
N{s)
D(s)
Jl
we must
=K
U(s)U(-s) ds
may
(11-39)
11-1.
Table 11-1
is
N(s)N(-s)
ds
D(s)D{-s)
= N-is-i + N- 2 s"- 2 +
+ Nts + N
= Dn s" + Dn-is"- + ... + DiS + D
.
IDoDt
_ N\D a + NlD 1
2
2D D D 1
2
j _ NjDoDj + (N - 2NoN1 )D D + NID 2 D
2DoD (-DoD + O1/J2)
2
/ = Nl(-DjD + DoD,D 2 + (N - 2N N )DqD D 4 + (N\ - 2NqN 2 )DoD
2DoD^-DoD\ - D\D* + DiD 2 D
j
D4
+ Nl(
-D\D\
-O2O3P4)
3)
Once
Uok (s)
is
determined,
if desired,
ok {s)
which
is
C(s)
(11-40)
R(s)
C(s)
k{s)
G(s)Uok (s)
(11-41)
u k{s)
W)
or
ok
= R-y ~GRT
(11-43)
Analytical Design /
Sec. 11.2
Equations
and
(1 1-38)
made more
(1 1-43) are
useful
579
by defining
(11-44)
G(s)
Y=
{k 2
k2
+ GG} +
DD + NN
DD
_ {k DD + NN} +
D
(11-45)
and
DD + NN}D
(11-46)
D
NR
DD + NN} + L{k DD + NN}-J +
(11-47)
Equation
(1
1-38)
now
U k
{k 2
_
~~
gives
{k 2
nk
which
k1
Example
N
R{k 2 DD
NR
(11-48)
is
finding
11-1
n
and the input transform
10
(11-49)
is
>-
(s)
(11-50)
transfer function
so that
(1)
Je
e 2 (t) dt
= minimum
(11-51)
< 2.5
(11-52)
(2)
f~ u\t) dt
J o
Since
10 and
D = s 2 we have
,
{k 2
To carry out
DD + NN}* = [k
s*
100} +
k 2 s*
100
we
(11-53)
write
= {a t s 2 + a,5 + Oo)(2* 2 - a lS + a
= als* - (a\ - 2a a )s 2 + a 2
(11-54)
Equating
like coefficients
on both
a\=k
(11-55)
$80
Chap. 11
=k
=
a2
a\
(11-56)
2a o
a\ = 100
(11-58)
(11-59)
(11-57)
which gives
aQ
10
we
(11-57),
get
= v 20fc
/
tf t
(11-60)
Therefore,
[k 2
DD + NN} + = (a
=
2s
+as+a
+ */20ks +
ks 2
(11-61)
10
and
{k 2
DD + NN}~ = to - */Wcs +
2
.{k 2
NR
DD + NN}-} +
10
(11-62)
is
10(0.5/.?)
- */20ks +
Iks 2
10_
(11-63)
0.5
at the origin
- 2
ok
Substituting
Uok
0.5
~ ks 2 + +/20ks +
~=
is
10
0.5.$
ks 2
+ */2Qk7+
i-L. ks
\0ks 2
^^=f
- */20ks
+
This integral
is
(11-64)
10
we have
0.5s
0.5s
s
+ >j20ks +
(11-65)
10
From Table
/.= N D + N D 2
2D D D 2
2
rry
-1,
(11-66)
where
N = 0.5, N = 0, D =
Q
= /20, D 2 = k. Thus
10, Z>,
0.125
S=
W20k
2.5
(11-67)
from which
k
The optimal
control
is
= 0.05
given by
tt s
UoKS>
0.5 s
- 0.05s 2 + s +
transfer function
(11-68)
10
is
M^=W)
Uo{s) =
'His)
0.05s
G(s)
200
;
+ 20s + 200
+S+
10
(11-69)
,.
'
Sec. 11.2
The
The performance index is the integral square error with constraint on the
control signal in the form of integral square control. No specifications are
given in the form of overshoot, bandwidth, rise time, gain margin, and
so on.
2.
final
system configuration.
It is interesting to note that the optimal system represented by the transfer function
of Eq. (11-69) has a damping ratio of 0.707 and a bandwidth of 14.14 rad/sec.
There is a great deal of flexibility in arriving at the final system configuration and
the transfer function of the controller. Figure 11 -2(a) shows the system with a series
controller. The transfer function of the series controller is determined as
c (s)
G(s)
(s)
- AfoCs)
(11-70)
20s
20
s
As an
is
20
H(s)
0.1s
G
and
lit)
f-
~\
e(0
u(t)
20s
.s
10
*-
s2
+ 20
c(t)
(a)
iiO
~\
e)
rA
"(O
10
+L
S^
0.1s
(b)
Fig. 11-2. (a) Analytically designed system in
controller, (b) Analytically designed system in
back
controller.
+-
c(t)
582
Example
Chap. 11
11-2
^)= ^i
The input
is
M 0) so
transfer function
that
D = s2
It is
Je
Then,
N = s
and
{k 2
The
(11-71)
DD + NN} + = [k
s*
s2
is
- s2 +
s*
1}
(11-72)
= (a2 s + a^ + a )(a 2 s 2 - a s +
= als* - (a\ - 2a 2 a )s + a 2
2
aQ )
(11-73)
Equating
like terms
on both
sides
=k
= */2k +
=
(11-74)
1
(11-75)
(11-76)
Thus
{k 2
DD + NN}+ = ks 2 + Jlk +
1,5
(11-77)
[k 2
DD + NN}- = ks ~ */2k
1,5+
(11-78)
and
\_{k
is
NR
-Or + 1)
_
DD + NNj'X LsVcs - ^/W+ls +
2
=:
1)J +
where we have considered that the pole at the origin is slightly to the
nary axis. The optimal control is now found from Eq. (1 1-47),
Uok =
~T
left
of the imagi-
s
ks 2
+ V2A: +
Is
(11-80)
1
Using Eq.
(11-39),
we have
= 0.378
transfer function
is
(1 1-48),
It is worth noting that the closed-loop transfer function contains the right-halfplane zero of G(s). This is a necessary condition for u(t) to be bounded. In other words,
if
(s) does not contain the term (s - 1) in its numerator in this problem, it
would
not be possible to satisfy the Ju
requirement. However, the analytical design
<\
formulation derived in this section naturally leads to the proper solution for this case.
It
should be pointed out that the analytical design can yield a physically
Parameter Optimization
Sec. 11.3
not
exist.
system
is
For instance,
known
well
it is
583
infinite if G(s) is
of the
form
<**
11 .3
(s
a)(s
t)(s
a> b C
'
c)
Parameter Optimization
The
analytical design
method introduced
in Section
1 1
.2 is restricted
to single-
input systems only. For multivariable systems, the spectral factorization problem
is
end
of the design
result
is
is
the
semi-
function.
As a
we may
consider a fixed-configura-
tion system but with free parameters that are to be optimized so that the fol-
As
satisfied:
Je
= C e \i)dt
Ju
= Cu\i)dt<K
J
(11-83)
we
(11-84)
when
/ == K,
and the
J
where k
system
is
is
=J +
e
k 2Ja
(11-85)
is
The
to be determined.
is
JU
=K
(11-86)
11-1, the
= J(Ku K ,...,K ,k
for
^=
i=l,
2,...,
that
is
is,
(11-87)
/ to be a minimum
performance index
is
ii
(11-88)
584
and the
Chap. 11
condition
sufficient
is
is
positive
definite
d 2J
d 2J
d 2J
"
3K dK2
'
d J
i
d 2J
_dKn dK,
Since d 2 JldK, dKj
d J
6K\
2
dK2 6K
V2/
dK
'
dK
'
dK
'
d 2J
dKMi
dK
d 2J
dKl
"
'
is
(11-89)
always symmetric. In
addition to the necessary and sufficient conditions of Eqs. (11-88) and (11-89),
the equality constraint of Eq. (1 1-86) must be satisfied.
illustrate
how
carried out.
Example
Consider the control system of Example 11-1. Let us use the cascade
controller configuration of Fig. 11-2, with the controller parameters
set as Ki and
2 The block diagram of the system is shown in Fig.
11-3. The problem is to determine the optimal values of
and K2 such that
11-3
minimum
(11-90)
2.5
(11-91)
J o
J=
{" u 2 (t)dt
J o
Kt)
.
to the system
is
r
v
u(t)
K lS
+ K2
c(t)
10
S2
Gp (s)
Gc (s)
The problem
From
is
=J +k
Fig. 11-3
Ju
2^7
[E(s)E(-s)
k 2 U(s)U{-s)] ds
(11-92)
we have
E(S)
=1
:
R(s)
0.5(s + K
+ K2 s + 10Ki
2)
+ G {s)Gp {s)
j2
(11-93)
and
U(s)
E(s)G c (s)
OSKiS
s2
+ K2 s +
10Ki
(11-94)
Substituting Eqs. (1 1-93) and (1 1-94) in Eq. (1 1-92), and using Parseval's theorem and
Table 11-1,
we have
Sec. 11.4
= 0.25(10*, +
Kl)
j_
lOKi^
2.5K1
k2
,v
585
(U "95)
lOK^z
Applying the necessary condition for J to be a minimum, we get from setting the
necessary conditions and the equality constraint,
dJ
dK
-0.25KJ
dJ
-L
dK2
Ju
The
last three
Kl
10AT,
K\
20K2
2.5
K =
=20
is
*?
(11-96)
lOA: 2
(11-97)
(11-98)
found
in
20
A:
Example
= 0.05
(11-99)
11-1. It
Ki equals
fact that
formulated a
one
=o
Hessian matrix
The
5k 2 Kj
Ki
which are
less flexible
problem
in this
is
K =K
problem by assigning
Then
also satisfied.
We
could have
would be only
would be quite
there
free
is
purely coincidental.
simple for the design criteria given in Eqs. (1 1-90) and (11-91), since the optimal value
of the parameter can be obtained by using only the constraint requirement of Eq.
(11-91).
11 .4
An
interesting
and
is
the design of a
linear feedback control system with specific eigenvalues. The system under
consideration may be represented by the block diagram of Fig. 11-4. The linear
process
where
is
The matrix
is
an
= Ax +
(11-100)
=-Gx + r
(11-101)
is
X n feedback
x =
is
matrix.
(A
BG)x
Br
Ax + Bu
(11-102)
586
Chap. 11
|AI-(A-BG)| =
(11-103)
~0
1
1
A=
a
The
reverse
is
-*-!
tf-2
a.
it is
is
To show
this,
we form
"
AB=
AB =
AB=
a.
_a\ - 2_
- a\
a2 a x
a 3j
Continuing with the matrix product through A" 'B, it will become apparent that
what the values of a,, a 2
a, are, the determinant of S =
[B AB A 2 B
A" _1 B] will always be equal to 1, since S is a triangular
matrix with Is on the main diagonal. Therefore, we have proved that if the
system is represented by the phase-variable canonical form, it is always state
regardless of
controllable.
G can be written
G = fci g 2
(11-104)
gn]
Then
1
1
BG
..
..
(11-105)
gi
a-i
g2
gn.
Sec. 11.4
587
The eigenvalues of A
|AI
(A
BG)|
= A"
=
Example
(11-106)
11-4
U(s)
s(s
g.
is
10
()_.
~
It is
l)(s
(11-107)
2)
desired to design a feedback controller with state feedback so that the eigenvalues
2,-1 +;'l,and 1 jl. Since the transfer function
By
the process
"o
is
-2
_0
L*3.
Since this
T*ii
x2
r~
x2
A3
(3
+g
A
like coefficients
(2
+g
G is of the form
(11-109)
3]
2 )X
is
+g,=0
(11-110)
it is
+4A 2 +6A +4 =
of the
g\
3 )k*
(11-108)
|_1_
-3J|_* 3 J
Equating the
G=te, g 2 g
The
is
last
(11-111)
g2
=4
gi
or
The
state
G=
[4
is
(11-112)
1]
shown
in Fig. 11-5.
Example
11-4.
Stabilizability*
588
Chap. 11
11.5
we
deter-
u(t)
-Gx(r)
(11-113)
r(t)
where
Er(t)
is
must
c{t)
u.
shows that
The output of
1-6
We
(11-114)
Gx(t)
Er{t)
Gx,(r)
is
(11-115)
exists.
is
The
closely
^X
\e
State-
observer
c(0
where x
is
an n-vector, u
is
=
=
Ax(?)
state observer.
time-invariant system
Bu(t)
Dx(/)
and
that
is
(11-116)
(11-117)
and c
is
Sec. 11.5
may
tives
T=
is
of rank
n.
We
(1 1-1 17),
show
shall
c(f),
we take the
and deriva-
(A')"
_1
(11-118)
D']
the input
derivative
c(0
589
is
(A') 2 D'
A'D'
[D'
u,
u{t),
on both
Dx(0
and
= DAxO) +
/.
We have
DBw(f)
(11-119)
or
t(t)
DBm(0
DAx(f)
(11-120)
Rearranging the
DBti(0
= DAx(0 = DA
x(0
+ DAB M (?)
(11-121)
equation yields
last
c(0
DBii(r)
= DA
DABm(0
x(/)
(1
1-122)
= DBm'"-
^)
- DABm<"- 2 >(0-
...
- DA-^BmO) = DA"-'x(0
D
DA
DA
c
c
c
DBh
- DBri
DABw
DB^"
'
(11-123)
DABm<"- 2
DAnl B
(11-124)
DA"
in terms of
c, u,
and
their
T=
=
must be of rank
[D
[D'
DA DA
A'D'
2
.
(A')
DA"
D'
]'
(A')"-'D']
n.
There are many ways of designing a state observer, and there is more than
one way of judging the closeness of x e (0 to x(f). Intuitively, the observer should
have the same state equations as the original system. However, the observer
should have u(t) and c(r) as inputs and should have the capability of minimizing the error between x(t) and x c (f). Since we cannot measure x(/) directly,
an alternative is to compare c(t) and c e (t), where
c c (0
is
Dx(r)
(H-125)
Based on the above arguments, a logical arrangement for the state observer
in Fig. 11-7. The state observer design is formulated as a feedback
shown
590
Chap. 11
"^
Xe
*e
\+
Ge
Fig. 11-7.
The
design objective
When
c e (0 equals
to select
c(/),
= Ax (?) +
xjj)
which
is
Bu(t)
(11-1 26)
is
with u and
c(f) as
resented by
.(0
(A
G.D)x.(0
*(0
G.c(0
is
1-127)
(1
shown
in Fig.
Since c(t) and x(f) are related through Eq. (11-117), the last equation
is
1-8.
also
written as
= Ax.(0 +
.(0
The
Bu(t)
is
+ G D[x(0 e
x,(01
(1
1-128)
observer will be identical to that of the original system. Therefore, the design of
the feedback matrix
tions to x(t)
and
There are
matrix
Section
is
many
possible
[x(f)
- ,(0] =
(A
- G D)[x(/) e
x.(0]
(1
1-129)
is
then
|AI-(A-G D)| =
e
Since
we
x e (t) as
(11-130)
may
Sec. 11.5
591
X)
Observer
State feedback
System
-^OH
HO
Fig. 11-8.
However,
x(t) rapidly.
the eigenvalues of
of matching
all
it
AGD
C
must be kept
A G^D may
in
mind
control only the denominator polynomial of the transfer relation, while the
numerator polynomial
is
(A
may
be established.
(A
BG)
given
and
has been
Thus, invoking the results of Section 1 1.4, the condition of arbitrary assignment
of the eigenvalues of (A
GD) or of (A'
D'G^) is that the pair [A', D'] be
completely controllable. This is equivalent to requiring that the pair [A, D] be
completely observable. Therefore, the observability of [A, D] ensures not only that
a state observer can be constructedfrom linear combinations of the output c, input
592
u,
and
Chap. 11
the derivatives of these, but also that the eigenvalues of the observer can
At
this point
an
may
example
illustrative
Example
11-5
is
W)
i(0
c(0
=
(s
IX*
= Ax(0 +
= Dx(0
(11-132)
2)
Bu(t)
(11-133)
(11-134)
where
o
.-2
-3_
D=
It is
B
0]
[2
are at
(A
G D) =
e
-10, -10.
G =
gel
(11-135)
LSel.
ple 11-5.
xi
O*io
for
Exam-
Sec. 11.5
593
is
+ 2g " ~
+ 2g e2 A + 3
(11-136)
= A 2 + (2g + 3)A + (6* + 2 + 2g. ) =
For the eigenvalues to be at A = 10, 10, the characteristic equation should be
|AI-(A-GJ!=
el
A2
Equating
like
+20A +
gel
state
(11-137)
The
100
=
=
8.5
23.5
is
shown
in Fig.
11-9.
As was mentioned
of x(0 and
responses of the observer will be identical to those of the original system, and the
feedback matrix G will have no effect on the responses of the observer whatsoever.
Figure 11-10
1.2.
xe i(t)
xl
gel
= 8.5
gel
= 23.5
0.8
0.6
*el(t)
(el=ge2=D
0.4
0.2
'
i
1
Time (seconds)
Fig. 11-10. State x\(f)
11-5.
states
594
Chap. 11
initial states:
x(0)
"0.5
*.(0)
Shown in the same figure is the response of x el (t) when the state observer is
designed for eigenvalues at A = -2.5, -2.5, in which caseg-,1 = 1 and^ 2
1. Since
the state observer dynamics are now slower, it is seen from Fig. 11-10 that the x (t)
ei
response deviates
Figure 11-11
design.
x 2 (t)
The
more from
the state
(t).
due to the fact that the elements of the feedback matrix, g el and g e2 appear in
the numerator of the transfer function relation between X2 (s) X {s) and x (0)
e2
2
- x e2 (0), but not in the relation between (s) eX (s) and x (0) - jcI (0). Taking
the Laplace transform on both sides of Eq. (11-129) and rearranging, we have
is
X(s)
- X(J) =
[si
(A
-G
D)]-'[x(0)
- Xc (0)]
(11-138)
where
s
[si
(A
-2-
- G D)]"
e
2ge2
2gelJ
(11-139)
|a-(A-G.D)|
Time (seconds)
Fig. 11-11. State
x 2 (t) and
in
Example
11-5.
Thus, we notice from Eq. (11-139) that selecting larger values for g el and ge2 to
numerator terms of the
e i(s) response. However, a larger value of g e2 will increase the gain factor between
2 (s) e2 (s) and x 2 (0) - x e2 (0). This explains the fact that in Fig. 11-11 the x e2 (t)
response for ge2 = 1 is a better observation of x 2 {t) than the one for g,, 2 =23.5.
give faster transient response for the observer will not affect the
X
X
Sec. 11.5
595
A more versatile design method for the observer may be effected by the use
A popular performance
J
The matrix
may be
is
\" {x
x.)'Q(x
(11-140)
x.) dt
states.
We
are
now
state-variable feedback.
is
shown
in Fig.
11-8.
From
Fig.
= Ax - BGx + BEr
i. = (A - DG - BG)x +
x
(11-141)
DG,x
BEr
(11-142)
diagram for the system is drawn as shown in Fig. 11-12 using Eqs.
and (11-142). The initial states are included on the state diagram, and
the branch gains are matrix quantities. It can be easily shown that the system
with the observer is uncontrollable (state) but observable. However, the lack of
controllability is in all the states x and x e The reason for this is simple; the main
objective is to control x while keeping x e close to x. It is not necessary to control
x and x e independently.
state
(11-141)
OXe (fJ)
= X e0
in Fig. 11-8.
596
From
Chap. 11
X(s)
These are
[si
(11-143)
e (s)
[si
(A
BG)]- BE/?(i)
]
+ (si - A)A-'x e0 +
G.DA-'x,,
(11-144)
where
5 2I
(1
1-145)
(1
1-146)
two systems
will
and
absent; that
means
that
is,
when
e.
More important
identical to that
is
is
the
the true state variables are fed back for control purposes. This
the
initial states
matrix
same
^ x,
Since, in general, x,
and subse-
quently of
due to x
to the
first
e,
as indicated
Example
11-6
As an example of
that has
the transfer
function
C(s)
U(s)
100
s(s
(11-147)
5)
= Ax(/) +
= Dx(r)
Bu(t)
(11-148)
(11-149)
where
"0
A=
-5_
Lo
D=
It is
[l
"
"
B=
LlOOJ
0]
= r-Gx
is
given by
(11-150)
where
G=fr, g 2
(11-151)
The eigenvalues of the closed-loop system should be located at; A = 7.07 y'7.07.
The corresponding damping ratio is 0.707, and the natural undamped frequency is
10 rad/sec.
Assuming
x and x 2 are
t
=1.
is
is
to
Sec. 11.5
can be
BG
completely controllable, so the eigenvalues of A
it is simple to show that the pair [A, D] is completely
The
and
u.
|Al-(A-BG)|
is
=0
(11-152)
(11-153)
or
A2
To
(5
lOOg^A
KXte,
=
K%! =
100^ 2
14.14
100
Thus gx
1 and g 2 = 0.0914.
The state diagram of the overall system, which includes the observer,
is
shown
in
Fig. 11-13.
rO
in
Example
11-6.
design of the observer may be carried out in the usual manner. The characequation of the observer is
The
teristic
|AI-(A-G.D)]
=0
(11-154)
where
G =
~ge\
(11-155)
i-gel-
598
Chap. 11
Let us assume that the eigenvalues of the observer must be at A = 50, 50.
These eigenvalues are of much greater magnitude than the real parts of the eigenvalues
of the system. Therefore, the transient of the observer due to the difference between
the initial states x and x c0 should decay very rapidly to zero. The corresponding
values of gei and g e2 are found to be
g.i
gel
= 95
= 2025
(11-156)
0.4
0.5
0.6
0.7
0.8
0.9
Time (second)
Fig. 11-14. Unit step responses of xi(t)
in
Example
11-6.
When x e o = x
^x
e0 ,
the response
= 10 and
is
10, which correspond tog ei =15 andg e2 = 25. The initial states of
x e0 = [0.5 0]'. Note that since the observer dynamics
are
much
is
also slower
Sec. 11.6
is still
greater.
599
from the
initial
*e i(0) =0.5.
It is interesting
and x ei (t). In
is
= 0,
G =
ideal response.
[15
illustrate
when x e i(0)
The case with ^(O) = *i(0) = is the
the importance of making x<, as close to x as
25]'.
0.25, 0.5,
and
0.8, with
These curves
possible.
1.2
x el =0.25
*10 ~*el0 ~
0.4
0.5
0.6
0.7
0.8
0.9
Time (second)
Fig. 11-15. Unit step responses of x\{t) of the system in
11.6
One of
the
applications
modern optimal
is
The
design
through
state feedback.
The
is
is
600
Chap.
are as follows
The
1.
solution
is
Can be used
2.
with multivariables.
Guarantees
3.
stability.
where
x(r) is the
is
= AxO) + Bu(
state vector
and
u(r) is the r
(1
control vector;
),
/
is
The design
x'C^Px^)
objective
" [x'(7)Qx(0
is
known
57)
A and B
and
B may
performance index
f
J H
is
1-1
u'(?)Ru(0] dt
(11-1 58)
as a quadratic performance
x(t f )
may
The
not be specified.
term under the integral sign of / represents a constraint on the
first
state variables.
unless
states, in
a diagonal matrix,
Q = 0.
For an nth-
Q may be defined as
"?,
...
?j
...
...
.0
0"
...
q1
Q=
The
is
which case
(11-159)
qn _
of
minimization of the performance index of Eq. (11-158) has the physical interpretation of keeping the states near the equilibrium state x = 0, while holding
the energy consumption to a reasonable level. For this design, the reference
is not considered and is set to zero, although the designed system may
be subject to nonzero inputs. The term "regulator" simply refers to the
input
still
Sec. 11.6
is
601
capable of
is
The
we
this section
may be
of
Principle of Optimality
Many
control strategy has the property that whatever the initial state
law of the
and
the control
respect to the state resulting from the control of the initial stages.
objective
is
from
x(t
given, in such a
way
=
is
fV(x,u)
(11-160)
dt
minimized. The state equations of the system may be that given in Eq.
Let
and
f (t
11-16.
<
f,
We
<
t ),
f
(1 1-157).
and
x^)
is
tx
< <
t
f , it is
expressed as
/,
Xx(?
f"F(x,u)A
(11-161)
-- *1
602
The
x0,) at
last
Chap. 11
x^)
u once
is
how
portion of the state transition, in order for the entire control process to be
/, the control for the time interval t
is
<t <tf
last section
of an
No
Hamilton-Jacobi Equation
In this section
we
derive the
is
design.
To show
linear systems,
we
is
f(x,u,?)
an
r-vector.
x(/)
where
x(t)
is
an n-vector and
The problem
is
u(() is
(11-162)
f"F(x,u,T)A
**
is
and non-
(11-163)
to
a minimum.
Let us denote the performance index over the time interval
<x<
/=JV(x,u,T)rfr
and
minimum
its
by
(11-164)
value by
S[x(t),
t]
Min /
Min
\" F(x, u, x) dx
J
(1
1-165)
(1
1-166)
Then
S[x(t
), t ]
Min J
Min
When
to
S[x(t f ), t f ]
+A
\" F(x, u, x) dx
J
and
t]
Then Eq.
Min
Q"
u, t)
dx
(11-1 67)
tr
+A<x<
S[x(t),
<x<t
(11-165)
F(x, u, t) dx
j" F(x,
u, x)
dx)
(1
1-168)
Sec. 11.6
603
j" +A
t]
Min
+ 5[x(f + A), +
F(x, u, t) dx
A])
(1
1-169)
(1
1-170)
(1
1-171)
S[x(t
A),
+ A] =
Min
Let
\"
J
F(x, u, t) dx
+A
t]
Min
(F(x, u,
0A +
S[x(t
A),t
A])
+e
(A)
where e (A)
is
The next step involves the expansion of S[x(t + A), f + A] into a Taylor
series. This would require the expansion of x(t + A) into a Taylor series about
A = first. Thus
+ A) = x(t) +
x(t
t]
^A +
(11-172)
written
is
Min(F(x,
u, t)\
+ (t)A +
S[x(t)
+ A]) +
e(A)
(1
1-173)
is
of the
S[x(t)
A],
S[x{t),
h(j), t
series as
follows
+ h(r), +
S[x(t)
A]
t]
^!k(() + *W>.f]
A
at
ox(.t)
(11-174)
higher-order terms
Therefore,
S[x(t
A),
= S[x(t) +
A]
^>
*>
f (x, u, f)A
ti
-Trf x
(
< -
Min(F(x,
'
where e^A)
>
u
'
** +
HM A
dJ}
(11-175)
f(x, u, t)
+ A]
higher-order terms
S[x(t),
u, r)A
we have
t]
'-
(11-176)
^^A) +
is
fl (A)
Since S[x(0,
t] is
factor
form
after cancellation
u{t),
Eq. (11-176)
is
sim-
- dS[
t]
f^'
Min
(F(x,
ii,
/)
+ ^gg^TXx, , /)) +
e 2 (A)
(1
1-177)
604
Chap.
As A approaches
which
known
is
becomes
^P
= Min F(x, u,
(
+ S^/X*, -, 0)
(1
1-178)
as Bellman's equation.
f.
^h
.
dSMt),t]dftx,v,t)
(11-179)
ft
du(t)
dx,(t)
dF(x,u,t)
dt(x, u,
Q <7S[x(t),
t]
__
""
du(t)
where
that
<?f(x, u, t)/du(t)
du(t)
(11-180)
dx(t)
t);
is.
df (x,
u,
'Mi
Mi
Ml'
du
dui
du
Mi
Mn
du 2
du 2
du 2
Mi
Mi
Ml
du r
du T
du r _
du(t)
t]
(Jacobian of
all
+ Fix", u, + dS
t]
t)
last
u, t).
dS[x\t),
which
is
m, u,
(1
1-182)
sum of the
H(x,
(11-181)
u(t) is
}$jJj
where
f)'
known
t]
+ H(x
o^
is
o
f)
known
also
as the
and
The
Hamiltonian
(11-183)
is
minimum
Now we
with the
initial
x(t)
condition, x(f)
Ax(r)
.
Bu(?)
(11-184)
is
= $x'(tf )Px(tf +
)
\" [x'(t)Qx(t)
u'(f)Ru(0] dt
(1
1-185)
Sec. 11.6
is
P and Q
and
R is
and
are symmetric
605
positive semidefinite,
definite.
minimum of/ be represented by 5[x(?), t]. Then, from the developlast section we know that the necessary condition of optimal
Let the
ment of the
control
is
must be
satisfied.
_ dS[x(t),t]
==
(11-186)
To
we
differentiate
result to zero
u(0
It
-R-'B' dy//^
ox
(11-187)
(t)
_o
Yrto
YfYA
_o,
rt
dS[x(t),
dS[x(t),
?]
V BR -, B
( dS[x(t),
t]
t]
>
dS[x(t),
Y BR _ 1R
t] \
dS[x(t),
t] \
(11-188)
The
last
equation
dS[x(t),t]
is
simplified to
_o,
rrt0x
o,
rt
'
dS[x(t),tT\
-i
W dS[x(t),t]
v
(11-189)
+ X WA
Since
dx(t)
Q is a symmetric matrix,
we
it
t] is
a quadratic
let
S[x(0,
t]
x'(t)K(t)x(t)
(1
1-190)
an n x n symmetric matrix.
control is determined in terms of K(t) by substituting Eq.
(11-190) in Eq. (11-187). We have
where K(/)
is
The optimal
u(0
-R-'B'K(0x(/)
(11-191)
last
^-x'(r)K(0BR- B'K(0x (0
|
(11-192)
(0
x'(0A'K(0x(0
x'(0[A'K(0
+ K(0A]x(0
(1
1-193)
606
Chap. 11
both sides of the equation represent scalar quantities. Thus, using Eq.
we have
since
dKjt)
- K(0A - A'K(f) - Q
K(0BR-B'K(0
dt
which
is
known
(11-194)
equation
is
K(t f )
=P
(11-195)
-G(r)x(0
(11-196)
G(0
R-'B'K(0
(11-197)
where
is
The
is
solved
from the Riccati equation of Eq. (11-194), with the boundary condition of
K(? r ) = P. The state equation of the optimal closed-loop system is given by
x(0
[A
BR-B'K(r)]x(0
^
J
is
shown
(1
1-198)
in Fig. 11-17.
X
x =
Ax + Bu
R-'B'K(f)
Fig. 11-17.
A few remarks may be made concerning the solution of the linear regulator
problem.
time
It
and nothing
(11-184).
regulator problem does not require that the process be observable, controllable,
or stable.
The
matrix
is
symmetric.
is
unknowns
Sec. 11.6
discussion
is, t
we have
607
differential Riccati
oo.
Now
let
= $ P [x'(0Qx(0 +
semidefiniteness
and
all
and
(1
terminal constraint
is
u'(/)Ru(01 dt
is
no longer
necessary.
1-199)
approaches
still
valid.
However, the
infinite-
1.
2.
The
process
is
any
solution of the linear regulator problem does not require that the
stable; in the finite-time case, controllability is
is
for finite
f,
J can
still
be
finite
even
if
an uncontrollable
is
state is unstable.
On
unbounded
[A', D'].
D]
approaches
is
may
be found in the
litera-
"
ture. 13 16
For the
set
becomes a
KBR^B'K - KA - A'K - Q =
(1
1-200)
A great deal of work has been done on the solution of the Riccati equation,
both for the
differential
form and
is
more
difficult to solve
than the
differ-
608
The
Chap. 11
differential Riccati
equation
is
methods:
1.
2.
3.
The direct numerical integration method simply involves the solution of the
nonlinear differential equations backward in time on a digital computer,
knowing K(^). The matrix
method
iterative
lim
K(f
is
A)
effected
by writing dK(t)/dt as
K(0
(11-201)
dt
backward
set
in time.
From
K(t)
y(0
and
and the
that
Mi)
t)
dx(t)
y(t);
(11-197), y(t)
(11-202)
is
= K(t)x(t)
(1 1-203),
K(t)x(t)
(11-203)
we have
K(t)x(t)
(1
y(0
which
is
is,
-Qx(?)
A'y(f)
1-204)
we have
(11-205)
Equations (11-198) and (11-205) are combined to form the canonical state
equations,
x(0"
BR'B'
-A'
x(0
(11-206)
Ly(0J
K(t f )x(h)
= P*^)
(1
(1
-207)
1-206) be
represented by
"
-BR-'B"
(11-208)
Sec. 11.6
The matrix
2.
If A, is
is
also an eigenvalue
that
W be the 2 X
=-X?
(11-209)
= W-'MW
(11-210)
is
Then
is,
where
609
1.
A,-
Let
M with
W be partitioned as follows
w
W w w
v
;
where W n W
W and W are n x n matrices. The transformed
Let
"
12
12 ,
21
22
(11-211)
respectively,
and are
state
and
related to the
x(0"
12
i(0*
(11-212)
lWV wV,
i(0.
(H-213)
"1(0"
= -A
i(0_
0" "1(0"
!
(11-214)
6 1 "a. i(0.
quation are
Since
(t)
y(
== e^'-'^itf )
e- A *-'.(t f )
(11-215)
(11-216)
it is
variable, t,
tr
Let
1(0
y(0
(11-217)
(11-218)
= e At^(0)
-A
$(t) = e 't(0)
5fc(T)
(11-219)
(11-220)
610
Since
Chap. 11
Eq.
(1 1-219),
and when
it is
(or
At
"e
(1 1-220),
"
Thus we rearrange
get
"X(T)
(11-221)
M)
e" At
f(T)_
we
Now
t(0)
(11-222)
Substitution of Eq. (11-222) into Eq. (11-221), the relation between (t)
is established:
and
X(t)
?(t)
- PW
-e~ A '(W 22
12
)-'(Wai
- PW M )r A i(t)
(11-223)
Let
H(t)
-e~ *(W 22
- PW 12 )-'(W 21 - PW n >-^
(11-224)
= H(t)X(t)
(11-225)
In order to return to the variables x(t) and y(0, we substitute Eq. (1 1-225) in Eq.
(11-212) and using Eqs. (11-217) and (11-218), we have the relationship between
Comparing the
last
[W 21
+ W^H^fW, + W I2 H(r)]-'x(r)
(1
1-226)
is
written
K(0
= [W 21 +
22
H( ?/
0][W
12 H(r,
t)Y
(11-227)
or
K(r,
t)
[W 21
+ W 22 H(t)][W u + W.jHCt)]-
(11-228)
problem obtained in
is
=
t
oo
~ t=oo
(1
1-229)
(1
1-230)
andEq.
lim H(t f
-0 =
(11-228) gives
K=
lim K(f f
t)
WjjWr,
(11-231)
Sec. 11.6
611
Although the above solution is of a very simple form, in general, to determine the matrices
and W, requires the solution of the eigenvalues and the
2
eigenvectors of the 2 x In matrix M. For high-order systems, a computer
solution becomes necessary.
we
shall first
consider the optimal linear regulator design of a first-order system and carry
out the design by pencil and paper. Then a higher-order system will be designed
with the aid of a digital computer.
Example
11-7
is
described by the
state equation
x(t)
with the
u(t)
initial
-2x(0
).
(11-232)
(/)
It is
such that
\xKtf)
f" t* (0
identify that
A = 2, B =
exponential method,
we form
M are X
P=
2=
* = min.
From
and
1,
(1
1-233)
R=
1.
M in Eq. (11-208),
~BR-
2 (0]
is finite.
I,
the matrix
M=
The eigenvalues of
1,
~-2
-r
_-l
2_
A2
= +/T.
B'~
(11-234)
-A'
Q
= V^T and
The eigenvectors of
M are
found to be
Pi
P2
(11-235)
_0.235_
Then
w
W = r^n
w
j\
_
1
xz
"
(11-236)
-4.:>35j
jo.:>35
2 2-
y%x
~|
_-4.235_
that they
It is important to note that the eigenvectors should be so arranged in
correspond to the requirement of the matrix T in Eq. (11-210). Substituting the ele= X 2 = */3~ in Eq. (11-224), we get
ments of W, P = 1, and
H(X)
The time-varying
Riccati gain
Kit,
is
-0.146<?" 2
We
Kit)
on
t,,x
(11-237)
T) =
=0,
Kit,)
+ 0.6189e"
- 0.146e- ^3'
= P = 1, which
2
0.235
1
condition
(1 1-228).
Thus
(11-238)
K(t,).
which
is
= oo. Thus
# = 0.235
is
(11-239)
'
612
Chap. 11
1.0
Steady-state
value of K(t)
0.52
K{t)
0.5
0.245
0.235
|
0.306
I
I
,
'/-I
0.5
t
t,
0.3
problem
is
= -R- B'K(t)x(t)
= -K(t)x(t)
<
l
infinite-time
problem
u(t)
(11-240)
tf
is
-0.23540
(11-241)
for the feedback system designed for the infinite final time
is
-0.235
Fig. 11-19.
Example
11-8
in
Example
11-7.
In
the pointing stability of the vehicle. In addition to the accuracy requirement, the
is
(/)
= Ax(/) + B(0
(11-242)
Sec.
1 1
.6
Control
moment
Vehicle
gyro
dynamics
Gimbal controller
u(t)
K,
K.s + K,
Fig. 11-20.
J* 2
LST
system.
where
"
H
J.
A=
J.
-Kj
-K
j
0"
B
IKj.
H = 600 ft-lb-sec
control
= 2.1 ft-lb-sec
K, = 9700 ft-lb/rad
Kp = 216 ft-lb/rad/sec
/ = 10 ft-lb/rad/sec
Jg
The design
objective
is
moment gyro
angular
momentum
gimbal inertia
gimbal-rate loop integral gain
gimbal-rate loop proportional gain
vehicle inertia
performance index:
J=\" [x'(0Qx(0 + Ru
(<)]
dt
(1
1-243)
614
Chap. 11
where
0"
Qi
1z
(11-244)
93
1iSince the states xi and x 2 represent the vehicle displacement and velocity, respec-
tively,
interest,
R=
more weights
We let q =
t
x 10 7 q 2
,
5000, q 3
</ 4
1,
and
l.
The optimal
control
is
given by
u\t)
where
The
-.R->B'Kx(0
(11-245)
(11-246)
digital
-2.75
-9700
G=
The
11-22.
-1.375
+ /2.33
[7071
5220
10.6
state
-1.375 -J2.33
is
0.99]
state feedback
(11-247)
is
shown
in Fig.
An input is added to the system through a gain of 7071. The latter is included so
that the output of the system will follow a step function input without steady-state
error. Figure
1-23
x 10" 8
(0)
rad.
The
initial
set to zero.
10.6
diagram of the
LST
is
a step
Sec. 11.7
Feedback
Design with
Partial State
6.00
7.50
9.00
7.50
9.00
615
0.05 r
0.00
3.00
1.50
4.50
Time
(sec)
0.08 r
o
X
0.041-
0.00
3.00
1.50
J_
_L
4.50
6.00
Time
Fig. 11-23.
input
11.7
is
(sec)
xi(t) of the
LST
a step function.
observable system to observe the state variables from the output variable.
However,
in general, the
system. Tt
is
is
is
It
gains from the accessible states, and no dynamics are involved in the controller.
Of
system
is
stabilizable with
Given the
if
the
H.0
Ax(f)
B(0
(1
1-248)
where
x(?)
u{t)
it is
is
= n x state vector
= scalar control
1
completely controllable.
m<
n),
y(?)
(m x
i,
and
y(0
Cx(?)
(11-249)
616
where
full
C is
Chap.
an
rank; that
X n matrix with
is,
the rank of
is
constant coefficients.
It is
assumed that
is
an
has
u(t)=-Fy(t)
where
(11-250)
= P [x'(0Qx(0 +
>
where
Ru*(t)] dt
(11-251)
to
The design objective is to find the optimal control u(t) such that / is minimized.
The reason for the choice of this performance index is that the design can be
associated with the linear regulator design.
Substitution of Eq. (11-249) in Eq. (11-250) yields
w(0=-FCx(f)
(11-252)
i/(0=-Gx(0
(11-253)
or
where
G = FC =
It
g2
fe I
...
(1
1 -25 1 ) for
+ G(joil -
1-253)
Q and R
A)"
>
is
if and only
if
(1
transfer
function
(si
of the
x(?)
and
A)-'Bf7(^)
system
(1
1-255)
co.
It is interesting to
The loop
some
(11-254)
g]
is
u(t) is
(1
G(sl
A) _1 B.
Thus
1-256)
1
is
relation of Eq. (9-74). In other words, the optimal system designed in the sense
The design
strategy here
is
(1
1-255) to select
G is then
considered as the optimal feedback gain corresponding to that Q. The approach
follows that of the inverse problem 11 of the linear regulator design, which
a weighting matrix
that
is
positive semidefinite.
The
resulting
involves the determination of the weighting matrices once the feedback matrix
is
known.
For convenience, we
R=
let
Sec.
1 1
Design with
.7
p(s)
si
(A
BG)
controllable,
is
The matrices
form.
"0
...
(11-258)
...
A is denoted by
= \sl-A\ = s* + as-~' + a^s"- +
characteristic equation of
q(s)
The
0..
2
1-257)
...
1
617
...
1
The
(1
written
-a,
Feedback
it
A and B are
Partial State
p(s)
= s +
(si
- A)'B
(a.
+ ^>"- +
]
...+
is
(a l
.+
a,
(11-259)
written
+g
(1
1-260)
Then
(11-261)
'?(')
and
G(sl
- A)
.g*s"
>B
g_
5- 2
1
...
g,
(11-262)
q(s)
Thus
[l+G(*I-A)-'B]=2g
In view of Eq. (11-255),
1
GC/col
we can
A)-'B|
[1
(11-263)
write
G(;coI
A)-'B][1
= p(Jco)p(-jco) >
G(-jcol
- A)~'B]
(1
1-264)
q(Jco)q(-jco)
Subtracting
from both
we have
qUco)q(-jco) -^ q
pUa>)p(JG>)
qUoj)q(-Jo>)
Let the numerator of the
last
(11-265)
equation be factored as
p(Joi)p(.-Jco)
- q(j(o)q(-j<o) = d(jco)d(-jco)
(11-266)
where
d(s)
Equation
(1
1-267)
is
a>"-'
of the (n
l)st
d^s"- 2
s"
a",
(11-267)
618
Chap. 11
d2
D=
(11-268)
- A)~'B
(11-269)
Then
D(sl
Equation (11-264)
[1
G(jcol
is
written as
- A)"B][1 +
G(-yo)I
= +
B'(-ycoI
G(jcoI
- AT'O'DOoI -
It
[1
^l
^-J^)
= +
q(joj)q(jco)
A)~'B]
Moore
12
that for
A)"'B
to be the optimal
(11-271)
A)"'B
we have
Q = DD
(11-272)
definite
in
mind
is
positive semi-
It
should be kept
does not
exist,
Example
11-9
= Ax(0 +
X')=* 2 (0
i(0
where
"or
L-i
It is
(H-270)
Ba(0
(11-273)
(11-274)
"0"
B=
o_
x 2 (t) through a
!""
J
is
The
is
of A
q(s)
and the
uHt)] dt
(11-275)
= Gx(t).
u(t)
characteristic equation
[x'(0Qx(0
is
|il- A|
=s +
2
(11-276)
\sl
(A
- BG)| = s + g 2 s +
2
from x 2
is
(11-277)
where
G=[0
g2 ]
(11-278)
Sec.
Design with
1 1 .7
Partial State
Feedback
619
= dz s + di
d(s)
D=
Then
(11-279)
d2 ]
[rf,
(11-280)
of the form
is
Q=
D'D
-d\
did
,
(11-281)
-d\di
From
Eq.
(1
-266)
p(s)p(-s)
we have
= s* + (2 - gi)s 2 + 1
= q(s)q(-s) + d(s)d(-s) =
s*
dl)s*
(2
+ d\ +
(11-282)
1
Therefore,
=0
d\
= gi
and
"0
0"
"0
0"
_0
d\\
Lo
gi\
(11-283)
Q must
solution does not exist for the problem given here. However,
exist.
It is
reasoned that
if
is
on the
state x 2
by assigning
which
is
the
the overshoot of
different values to
d2 and
,
observing the effect on the roots of the characteristic equation of the feedback system.
d\
d\
d\
that
=2:
=4:
p(s)
p(s)
p(s)
=s +s+1
= s 2 + *fls
= s + 2s + 1
2
damping ratio C
damping ratio
damping ratio
= 0.5
= 0.707
= 1.0
last
in Eq. (11-258)
it is
logical to
absorb these coefficients in the feedback matrix G. Then the modified matrix
A would
be
"0
...
0"
A*
(11-284)
0...
...
The
state equations
= A*x(r) +
x(/)
and the
state
feedback control
Bu(t)
(11-285)
is
u(t)
-G*x(r)
(11-286)
620
Chap. 11
where
= G-[0
= igi i
G*
It is
1]A
...
A*
gi
q(s)
(11-287)
a2
a]
will
always be
(11-288)
s"
= Ax(0 + B(0
(0
where
A=
-3.
[g
[x'(0Qx(0
From
Jf~
to
_1_
G=
We
is
B =
-2
"0"
0"
0]
2 (01 dt
G*
fc,
.vi
is
3].
=
=
\sl
- BG)| =
+ 2s + g
(A
3.s
\sl
(A*
is
- BG*)|
(11-289)
Let
d(s)
From
= d S 2 + d2 S + d,
(11-290)
we have
Eq. (11-266),
p(s)p(-s)
(11-291)
d\
d\=g\
d\ = 5
(11-292)
(11-293)
~dl=6g -4
2did3
we have
(11-294)
d\
1.5278^
(11-295)
From
<
1.5278
It is interesting
teristic
is
2.618
(11-296)
Eq. (11-292),
g,
Xi
< 2.618
(11-297)
Routh-Hurwitz
equation of Eq. (11-289) shows that the closed-loop system with feedback from
stable for #i
< 6.
Chap.
References
D'D
d\
didz
d^df
did2
d\
d2 d3
-did 3
dzdi
d\
two values of ^i
is
621
the
(11-298)
_
REFERENCES
Analytical Design
1
2.
4.
M. Brasch and
tors,"
5.
E.
J.
IEEE Trans.
Linear
1967.
J. B. Pearson, "Pole Placement Using Dynamic CompensaAutomatic Control, Vol. AC-15, pp. 34-43, Feb. 1970.
Feedback,"
6.
7.
E. J. Davison and R. Chatterjee, "A Note on Pole Assignment in Linear Systems with Incomplete State Feedback," IEEE Trans. Automatic Control, Vol.
AC-16, pp. 98-99, Feb. 1971.
J.
C. Willems and
tion,
S. K. Mitter, "Controllability, Observability, Pole Allocaand State Reconstruction," IEEE Trans. Automatic Control, Vol. AC-16,
IEEE
Trans.
Trans.
Design of Observers
9.
10.
11.
12.
D. G. Luenberger, "Observers for Multivariate Systems," IEEE Trans. Automatic Control, Vol. AC-11, pp. 190-197, Apr. 1966.
D. G. Luenberger, "An Introduction to Observers," IEEE Trans. Automatic
Control, Vol. AC-16, pp. 596-602, Dec. 1971.
B. D. O.
Inc.,
Anderson and
Englewood
J.
Cliffs, N.J.,
B.
1971.
Control, Prentice-Hall,
622
Chap.
14.
15.
Is
ASME,
M. Athans and
New
16.
"When
R. E. Kalman,
J.
York, 1966.
Inc.,
Englewood
Cliffs,
N.J., 1968.
PROBLEMS
11.1.
e\t) criterion,"
f"
e\t) dt
can be expressed as
F(s)F(-s) ds
*'
J -j.
= dE(s)/ds.
where F(s)
11.2.
= -
A unity-feedback control
r(t)
/\ /\
system
is
in Fig.
PI 1-2.
e(t)
\/
shown
s(s
c(t)
+ a)
Figure Pll-2.
(a)
J=
(c)
(d)
11.3.
The
e 2 (t)\dt
i
(b)
transfer function of
linear process
U(s)
is
s2
Problems
Chap. 11
623
KO ~
c(t)]
= minimum
dt
-i
dt<l
The
11.4.
A linear process is
is
described by
x2
= x2
=u
is
the control. It
'xt
is
desired to mini-
] rfc
For the control system shown in Fig. PI 1-5, the input r{t) is a unit step function.
Figure Pll-5.
(a)
K2
Determine Ki and
subject to
Ju
minimized,
is assumed to be zero.
and the values of K and K2 are as determined
in part (a),
f~
e\t) dt
/(/)
= 0,
find the
maximum
["
d 2 (t)
dt
Je<\
(d)
Ju
(b) If
(c)
=
The
Je
is
<
Ju
<
Nd(f) that
criteria:
minimum
Repeat part (b), and find the maximum strength of the impulse disturbance
is less than or equal to the minimum value found in (c).
N such that /
624
11.6.
The
The
Chap. 11
control system
shown
in Fig.
reference input
Je
J"
is
J"
and
e 1 {t) dt
" 2(/) dt
-25
minimum
'
*~c(t)
Figure Pll-6.
11.7.
x(0
c(t)
= Ax(0 + B(0
= Dx(r)
where
"1
0"
Lo
0_
"1"
B=
D=
L
[2
-U
11.8.
with the
J
where
and
tf
i J\"[2x\t)
to
uHt)]dt
sec.
11.9.
11.10.
oo.
= 0.5x(t) +
u(t)
rr
T
(\e~<x 2
e~'u 2 ) dt
u(t) that
minimizes
Problems
Chap. 11
Show
is
625
given by
11.11.
_ ptp-T
Draw
i.
X\
2x lXz
4x\
u 1 ) dt
a block diagram for the closed-loop system. Find the damping ratio of
the system.
11.12.
(0
= Ax(/) + B(0
where
"0
A=
l_4
It is
B =
-4
1J
1.
G = [#i
g 3 ] such
that the
performance index
[x'(0Qx(0
u 2 (t)] dt
= minimum
to
Formulate the problem using the method described by Eqs. (11-284) through
(11-288). Find the bounds on g^ and g 3 such that solutions to the optimal
linear regulator problem with the specific partial state feedback exist.
APPENDIX
Frequency-Domain
The
Plots
is
is obtained by setting
in G(s)H(s). In control systems studies, frequency-domain plots of the
open-loop transfer function G(jco)H(jco) are made for the purpose of analysis of
= jco
\G(jco)\
/GUa>)
(A-l)
where G(jco) denotes the magnitude of G(jco) and /G(jco) is the phase of G(jco).
The following forms of frequency-domain plots of G(jco) [or of G(jco)H(jco)]
versus co are most useful in the analysis and design of feedback control systems
in the frequency domain.
|
1.
2.
Polar plot: a plot of the magnitude versus phase in the polar coordinates as co is varied from zero to infinity.
Bode plot (corner plot): a plot of the magnitude in decibels versus
co (or
3.
log 10
co) in
A.1
A.I
627
The polar
is
From
is
varied
from zero
to
ping of the positive half of the imaginary axis of the .s-plane onto the plane of the
function G(jco).
frequency
co
x ,
i/CO
s-plane
./co 2
-jOJ l
,,/ImC
G(/co)-plane
*-ReG
628
Frequency-Domain
Plots
App.
is
referred to as positive,
and
clockwise as negative.
To
G(s)
where
T is
(A-2)
+Ts
+jcoT
a positive constant.
Putting s
= jco,
we have
G(jco)
last
expression
G(jco)
*J\
+ <o r
2
(A-3)
is
written
/-tan-' coT
(A-4)
When a> is zero, the magnitude of G(jco) is unity, and the phase of G(jco) is
at 0. Thus, at co
0, G(jco) is represented by a phasor of unit length directed in
the 0 direction. As co increases, the magnitude of G(jco) decreases, and the phase
becomes more negative. As co increases, the length of the phasor in the polar
coordinates decreases, and the phasor rotates in the clockwise (negative) direction. When co approaches infinity, the magnitude of G(jco) becomes zero, and
the phase reaches 90. This is often represented by a phasor with an infinitesimally small length directed along the 90 axis in the G{ j'co)-plane. By substituting other finite values of co into Eq. (A-4), the exact plot of G(jco) turns out
to be a semicircle, as
As
shown
in Fig. A-2.
= + jcoT
1
G(jco)
where 7^ and
T2
+jcoT
(A-5)
/ImC
is
written
G(/'co)-plane
ReC
V
Fig. A-2. Polar plot of G(jco)
1/(1
w2 T 2
+jcoT).
A.1
G(jco)
The polar
T2 and r,
= yy
+
or
co
/tan"
T]
coT 2
unity as co
If
is
T2
is
the phase
is
T2
is
tan~' coT
upon the
(A-6)
magnitudes of
relative
is
629
posi-
tive. If
less
less
/ImG
G(/co)-plane
co
T2
*-ReG
> T,
(T2
,,/ImG
G(/co)-plane
zyr,
+-ReG
(T2
<T
It is
tion
is
(1
apparent that the accurate plotting of the polar plot of a transfer funcis of high
the final figure of the polar plot, for a wide class of transfer functions. However,
it is
be completely
is
630
Frequency-Domain
App.
Plots
facilitated
is
Example A-l
make a rough
Substituting s
jco in Eq. (A-7), the magnitude
oo are computed as follows:
co
at co
and
lim G(jco)
= lim ^ =
= lim /-) =
lim IG(jco)
lim G(jco)
|
lim
<0-oo
lim /G(jco)
we determine
oo
lim
-^
if
-90
(A-9)
(A-10)
CO
A^L =
plot of G(jco) at CO
the intersections,
(A-8)
-180
and
(A-ll)
co
oo are ascertained.
G(y'C0)-plane.
If the polar plot of G(yCO) intersects the real axis, at the point of intersection, the
is
zero; that
is,
Im[G(/co)]
(A-l 2)
Re[G(yco)]
+ j Im[G(/co)]
(A-13)
we must
by multiplying its numerator and denominator by the comdenominator. Therefore, G(y'co) is written
rationalize G(jco)
plex conjugate of
C(im\
KJ
'
its
10(-yco)(-/CO + 1)
+ l)(-jco)(-jco
jcoO'co
-10C0
_
~ co* + co
1)
_
2
10C0
co*
co 2
,.
lA" 14
'
which gives
mGUco)]= (o(
(A-15)
l)
and
Re[G(7co)]
When we
set
Im[G(yco)] to zero,
we
^^L
get CO
oo,
(A-16)
meaning
that the G(jco) plot has with the real axis of the plane
is
at the origin.
Similarly, the intersection of the polar plot of G(jco) with the imaginary axis is
found by setting Re[G(yco)] of Eq. (A-16) to zero. The only real solution for co is also
co = co, which corresponds to the origin of the G(/co)-plane. The conclusion is that the
polar plot of G(yco) does not intersect any one of the two axes at any finite nonzero
A.1
631
l/lmG
G(/co)-plane
--Re
upon
frequency. Based
at co
and
co
Example A-2
this
co
1)].
10
made
make a rough
'
s(s
l)(.s
(A-17)
+ 2)
at co
and
oo:
lim GC/eo)
|
HJ-.0
lim /Gt/c)
Km
lim 14JCO
G(yco)
-90
(A-19)
= lim -g =
= lim
/^j
we
(A-18)
CO
t-0 /
lim /GC/OJ)
G(ya>)-plane,
lim
cu->0
OJ-.0
To
desired to
tions are
I0l[s(s
G(s)
it is
(A-20)
-270
(A-21)
G(uiC0 )
+ 1)(-/cq + 2)
+ 2)(-jco)(-jco + IK-jco +
m-jco)(-jco
=
jcoUco
'
l)(jco
is
9co*
(A " 22)
written
-30a> 2
2)
co 2 (2
- co
- a)>)
+ co 2 (2 - co
JlOaKZ
2 2
)
9co*
2 2
)
(A-23)
632
Frequency-Domain
App.
Plots
We set
Re[G(/co)]
-30
9co 2
(2
- co
(A-24)
2 2
)
and
Im[G(/w)]
Equation (A-24)
is
-10(2
9 3 +
satisfied
co 2 )
- CO
(A-25)
1 2
)
when
satisfied
CO
tion (A-25)
co(2
oo
intersects the
when
CO 2
=2
which gives the intersection on the real axis of the G(/C0)-plane when co = *J~2
rad/sec. Substituting co = /2~ into Eq. (A-23) gives the point of intersection at
G(yV2)=-j
(A-26)
it
is
i/IraC
G(/oo)-plane
Re G
= lOMs +
l)(s
2)].
Bode
functions that
is
633
the real and imaginary axes in the transmay again be quite involved. Further-
manipulation
it is
somewhat awkward
for design
purposes. We shall show in the next section that approximate information on the polar
plot can always be obtained from the Bode plot, which is usually sketched without any
computations. Thus, for more complex transfer functions, other than using the digital
computer, sketches of the polar plots are preferably obtained with the help of the
Bode
A.2
Bode
plots.
The
corner plot,
is
terms of
is
its
basically
an approximation method in
is approximated by
Bode
1.
The magnitude
plot of the
Bode
plots of
most
may be approximated by
straight-line
segments. This makes the construction of the Bode plot very simple.
Since the corner plot
is
may be
plot,
which
is
In general,
it
we may
0(5)
+
^+^
+
+
s-(s
where
As an
is
pj(s
*>/
p2 )
alternative, the
G{S)
s'(l
+ *->
+ p)
(A-27)
{'
(s
is
written
Ta sKl+Tb s)---(l+Ts)
(A 28)
where A" is a real constant, and the Ts may be real or complex numbers.
In Chapter 8, Eq. (A-27) is the preferred form for root loci construction.
However, for Bode plots, the transfer function should first be written in the form
634
Frequency-Domain
Plots
App A
of Eq. (A-28). Since practically all the terms in Eq. (A-28) are of the same form,
without loss of generality, we can use the following transfer function to illus-
Bode diagram
+T
s(l
T T 2 Ta
where K,
t ,
order polynomial,
f,
and
ju
2^/j.
G(jco) |
/i
20 log,
/G(jco)
//
co/co n ,
in decibels
is
is
by 20; we have
20 log, o
ju
= 20 log I0 K + 20 log,
+ jcoT
(A-30)
+ j(oT - 20 log, ja>
+ joaT.\ 20 log 10 + J2{ M - y}\
20 log.o G(jco)
+
-
+ j2{/u -
s)(l
written
jjco
many more
may
construction of the
in general, G(jco)
Bode
may
plot
We
Constant factor
2.
(jco)
3.
4.
Complex
that,
(1
poles or zeros
(1
+ jcoT) q
+ j2/z
2 r
fi )
We
shall
now
Bode
types of factors.
Constant Term,
Since
KiB =
20 log I0
K=
constant
(A-32)
and
W
(0
K>
*<o
(A " 33)
Bode
A.2
20 log 10 K
Magnitude
635
dB
0dBf_
indB
Function
90
ArgK
Phase in
degrees
-90 -
- 180
0.1
100
10
co rad/sec
Fig. A-6.
Bode
the
Bode
plot of constant K.
shown
is
in Fig.
A-6
in semilog coordi-
nates.
|Oco)
'|=
given by
is
20/7 log,
dB
co
(A-34)
> 0. The last expression for a given p represents the equation of a straight
ctJ
line in either
lines
in decibels
may
is,
dlog ioQ)
These
( 2P
lo Sio ca)
lines all
co in
variation in
is,
nates.
co,
that
co
dB
(A-35)
in
from
dB
20/7
equivalent to
decade of
is
octave
and
if CO2/CO1
co 2 is
2.
given by
number of decades
io Bio (a>i/<i)
og
log,o 10
Similarly, the
number of octaves
logi
co 2
and
number of octaves
0.301
is
0.301
("h)
\coJ
(A-36)
co, is
W<o,) =
log,
log
(coA
\coJ
(A . 3?)
given by
decades
(A-38)
636
Frequency-Domain
Plots
App.
we have
20/7 dB/decade
2Qp x
0.301
6p dB/octave
0-dB axis
is
at cu
at s
=--
20 dB/decade, and
0,
(A-39)
the magnitude
1.
is
written
l(jco)
= p
90
(A-40)
(jco)
are sketched as
0Q
T3
s:
<4-
100
rad/sec
cj
Fig. A-7.
Simple Zero
(1
-\-
Bode
plots of Uo>) p
j<oT)
Let
G(jco)
where
T is
a real constant.
I
G(ja>) |
= + jaT
The magnitude of
20 log
(A-41)
G(j(o) in decibels
is
written
( A-42)
A.2
Bode
637
To
since co
is
At very
co
T2
G(jco) dB
|
20 log
G(jco)
~ 20 log
o 1
dB
(A-43)
neglected
co
T2
by
G(jco)
U = 20 log
I0
G(joi)
~ 20 log
=
I0
Jco 2 T 2
(A-44)
20 log, coT
co
-=-
(A-45)
approximate magni-
40
II
II III
I
I
Ml
30
20
10
10
-20
30_LL
40
0.01
Fig. A-8.
and G(s)
0.1
1/(1
Ts).
Ts
638
Frequency-Domain
Plots
App A
Table A-l
>T
\osiqC0T
|1
+ jwT\
\l
+ ja>T\
/l
+ jwT
(dB)
0.01
-2
0.1
-1
1.04
0.5
-0.3
-0.12
0.76
1.0
0.5
5.7
0.043
1.12
26.6
1.26
37.4
1.41
45.0
1.31
0.117
1.65
4.3
52.7
1.73
0.238
2.0
6.0
60.0
2.0
0.3
2.23
7.0
63.4
5.0
0.7
5.1
14.2
78.7
0.0
1.0
10.4
20.3
84.3
Table A-2
Straight-Line
Approximation of
+ j<oT\
coT
(dB)
decade
below corner
0.1 (1
|i+;an
Error
(dBY
{dB)
0.043
0.043
frequency)
0.5 (1 octave
1.0
below corner
frequency)
0.76
.0 (at the
corner
frequency)
1.31
4.3
2.3
2.0 (1 octave
above corner
frequency)
10 (1 decade
20
20.043
0.043
above corner
frequency)
Table A-2 gives a comparison of the actual values with the straight-line approximations at some significant frequencies.
The
totes
it is
at
is
error between the actual magnitude curve and the straight-line asympsymmetrical with respect to the corner frequency \/T. Furthermore,
useful to
remember
is
dB
and
dB
decade above and below the corner frequency, the error is dropped to approximately 0.3 dB. From these facts, the procedure in obtaining the magnitude curve
of the plot of the first-order factor (1 + jcoT) is outlined as follows:
Bode
A.2
639
2.
3.
l/T.
dB, with the two lines interesecting at co
line at
locating the
obtained
by
actual
magnitude
curve
is
If necessary, the
1.
line
Usually, a
The phase of
G(jco)
+ jcoT
is
/G(jco)
written as
tan"'
mT
(A-46)
As shown
maximum
is
less
deviation of the
than
6.
Table A-l
Fig. A-9.
G(s)
Simple Pole.
1/(1
11(1
Ts and
Ts).
+ jcoT)
When
1
G(Ja>)
1
the magnitude,
G(jco)
in decibels
it is
Bode
simple to extend
<
coT >
coT
is
is
all
We
|
can write
G(jco) dB
|
(A-47)
+jcoT
G(jco)
U=
dB
-20 log 10
(A-48)
coT
(A-49)
Thus the corner frequency of the Bode plot of Eq. (A-47) is still at co = l/T.
At high frequencies, the slope of the straight-line approximation is +20 dB/
decade. The phase of G(jco) is 0 when co =
and 90 when co approaches
640
Frequency-Domain
Plots
App.
The magnitude and the phase of the Bode of Eq. (A-47) are shown in
A-8 and A-9, respectively. The data in Tables A-l and A-2 are still useful
infinity.
Figs.
made
For
instance,
to the numbers.
is
dB.
,*
=
1
2C,.,
+ a*
(2/co)s
(A50)
+ (l/cofis*
1,
would
have two unequal real poles, and the Bode plot can be determined by considering G(s) as the product of two transfer functions each having a simple pole.
Letting s
= jco,
The magnitude of
1
G(jco) dB
|
is
frequencies, co/co
(A" 51)
[1-W1+W.J
G(j(o) in decibels
20 log 10 G(jco)
At very low
<C
= 20 log 10
Eq. (A-52)
G(jco)
=5
(gJJ
may
-20
4C
(A-52)
(g)
be written as
log 10
dB
(A-53)
Thus the low-frequency asymptote of the magnitude plot of Eq. (A- 50) is a
on the 0-dB axis of the Bode plot coordinates.
At very high frequencies, eo/con ^> 1 the magnitude in decibels of G(jco) in
Eq. (A-50) becomes
straight line that lies
\G(jco)U
20 log 10 \GU<o)\
= -20 log
10
^/g-)
(A-54)
dB
401og,f-^\
This equation represents the equation of a straight
is
line
with a slope of
intersection of the
two asymp-
-401og 10 (^-)
= 0dB
(A-55)
\co n /
get
co
(A-56)
co
co
, is considered to be the corner frequency of the
second-order transfer function of Eq. (A-50), with the condition that
1.
The actual magnitude plot of G(jco) in this case may differ strikingly
<
lines.
The reason
for this
is
and phase
Bode
A. 2
641
curves of the G(ja>) of Eq. (A-50) depend not only on the corner frequency co,
but also on the damping ratio . The actual and the asymptotic magnitude plots
the
of G(jco) are shown in Fig. A-10 for several values of . The errors between
s. The standard
set
of
same
for
the
1
A-l
in
Fig.
shown
are
of
curves
sets
two
40
1
Mil
Mill
mi
II Mil
30
f
20
10
0.5
= 0.05
/ al
10
= 0.707
1.0
-20 -
-30 -40
in
0.1
0.01
II III
iiN
u = cj/w
Fig.
25/o))
Bode
plot of G(s)
1/[1
(sl<oY).
25
20
15
10
-5
-
10
15
llll
0.1
0.01
2.rM,)
(s/co)*].
Mil
100
10
1.0
642
Frequency-Domain
Plots
App.
procedure of constructing the magnitude portion of the Bode plot of a secondorder transfer function of the form of Eq. (A-50) is to first locate the
corner
frequency
making corrections
f.
The phase of
G(jco)
given by
is
plotted as
is
shown
0
1
A- 12
in Fig.
W4;
-45
f=1 '//V*H
-90
Mil!
II III!
f = 0.05
t^/0.1
0.6//
0.4
(A-57)
0.2
0.3
135
180
MM
0.01
1
1
0.1
+
The
INN
MM
100
1/[1
---
2(s/)
(*/*)*].
analysis of the
Eq. (A-50)
w/w
u =
Fig. A-12.
1.0
may be applied
transfer function of
two complex
zeros. If
G(s)=
+%s
co
the
Bode
plot of G(jco)
may be
(A-58)
(On
As an
Example A-3
illustrative
let
us consider
G(s)
The
first
jco.
step
is
10)
2)(j
10(5
s(s
(A-59)
5)
GUco)
10(1 +yo.ico)
jco{\ +y0.5(U)(l +/0.26t>)
set j
(A-60)
This equation shows that G(jco) has corner frequencies at co = 10, 2, and 5
The pole at the origin gives a magnitude curve that is a straight line with a
slope of 20 dB/decade and passing through the co = 1 rad/sec point on the co axis at
rad/sec.
dB. The total Bode plots of the magnitude and phase of G(jco) are obtained by
Magnitude-Versus-Phase Plot
A.3
643
Fig. A-13.
Bode
plot of G(s)
[10(j
10)]/[4f
2)(j
5)].
adding the component curves together, point by point, as shown in Fig. A-13. The
actual magnitude curve may be obtained by considering the errors of the asymptotic
curves at the significant frequencies. However, in practice, the accuracy of the asymptotic lines is deemed adequate for transfer functions with only real poles and zeros.
A.3
Magnitude-Versus-Phase Plot
is a plot of the magnitude of the transfer
in
degrees, with co as a parameter on the
phase
function in decibels versus
applications
important
of this plot is that it can be supermost
curve. One of the
posed on the Nichols chart (see Chapter 9) to give information on the relative
stability and the frequency response of the closed-loop system. When the gain
of the transfer function is varied, the plot is simply raised or lowered
factor
of adding the individual plots for cascaded terms in the Bode plot does not
carry over to this case.
magnitude-versus-phase plot
is
in obtaining the
digital computer is used to generate the data. Usually, the magnitude-versusphase plots are obtained by first making the Bode plot, and then transferring the
As an illustrative example, the Bode plot, the polar plot, and the magnitudeversus-phase plot of the function
G(s)
IOC?
s(s
10)
2)(s
(A-61)
5)
644
'
Phase
crossover/^'
App.
Im
_^^
G(s)-plane
3
<5
co -> oo
v?/
^y
'
Jl
Re
*
\V^.
Gain
crossover
1000
(a)
G(/u)
pa
a
CO
10
3~
\y
J*
^r co =
o.
30
20
co
'
rad/sec
Gain crossover
10
co= 10
Phase
crossover
20
10
-270
*/
= 5.5
to
,
-220
-180
-90
-140
/g(/co)
(c)
diagram,
are sketched as
(c)
ri0(.s
10)]/[4r
2)(j
5)]. (a)
Magnitude-versus-phase plot.
shown in
Fig. A-14.
The
The relationships among these three plots are easily established by comparing the curves in Fig. A-14 without the need of detailed explanation. However,
for the purpose of analysis and design, it is convenient to define the following
terms
Gain-crossover frequency. This
the transfer function G(jco)
is
is
dB. The following interpretations are made with respect to the three types of
plots:
Polar Plot.
A-14(a)].
(or points)
is
where G(jco)
[
[Fig.
A.
Bode
Plot.
(or points)
is
is
80, or
Bode
is
where
A- 14(c)].
180.
Polar Plot.
is 1
is
645
(or points)
is
is
where
APPENDIX
Laplace Transform
Time Function
F(s)
fit)
_1_
s*
t" (n
= positive integer)
e -at
e~<"
Ct
a)(s
e -6 '
6-a
b)
co 2
<&n
j2
2Zcon s
(i
+ sry
r(-i)!
7tB 2 e-'/r
0i n 2
(1
646
Ts)(s*
2Zcos
27to
co2)
where
TW
tan-
'
<n-1 r -rr
V(l-
^i."
)d
(eWl - ^t
- 27ca -
rW
0)
App. B
Laplace Transform
Time Function
F(s)
fit)
(On
sin cot
+ (On 2 )
{s 2
^m"
(1
C0n 2 )
where
vi+rw
'
a> 2 )
where
tan -1
s(l
(0 2 )
e-'/r
Ts)
-t-^-e-'i?
Ts) 2
!
1
rw
2rCco
7"
(On 2
S(\
Ts)(.S 2
e -c<,( sin
2liC0S
Vi -
(On )
where
"/1
tan"'
2Ccos
co 2
_r
5 2 (1
co 2 (l
<f>
S2
2{C0n S
co 2 (l 4- 05)
(s 2
(On 2 )
^(On
where
co\/l
5 "1
2t
4>)
rw)
Wl ~ ^
+ T 2(On 2
-4
sin
tan"'
a 2 co 2
rc<
f2 '
)d
tan"'
(t
e -e*r gin
a(On
acoWl
_ ( 2 t _ 0)
- 2i(0nT + T 2co 2 )
r Wl ~^
f
1 - TcO n C
(q,^ i
- 2T+
as)
+ C0 2
Tc
eW(l V1 ~ ^ +
2C0)r
2 tan"
(On^jl
2crco
e-
Ts) 2
+ (On 2 )
v' 1
(gW
d -
tan-
where
s 2 (l
T*COn 2
2f
00
(On 2
2r.(0S
Ts)(s 2
C"
0-2 tan"
where
"
t
i + ^^/r^"'"'
'
(On 2
s 2 (s 2
<K\
cos (0t
i(l
C0 2
(m
vl-( z
+
2Zcos
s(,s 2
lin
<f>
(On 2
s(s 2
r~' IT
2C
+ T 2(On
-1
= tan con T
(On
Ts)(s 2
2T)e-<lT
(aWT^P/ +
<f>)
~^
af.(o
sin (cot
0)
where
<6
tan -1 aa>
648
App. B
Laplace Transform
Time Function
F(s)
fit)
coHl
(1
Ts)is 2
+ co 2
+ a 2co 2
+ t 2 cb 2
2gfra
2rcco
where
o 2 )
where
as)
2Cws
<j>
"J^g? -
tan-
-1
tan
VI
-2aCco
co 2 (l
as)
Ts)(s 2
co 2 )
where
ca 2 (l
j(1
J*)(i 2
co 2 )
tf
T)
tan -1 a>r
2Coa>B
sin
tan"'
2 )(l
+ a 2e 2
2TCco
+ r 2 <o ^
(o^rqp/ + + r
+as
s 2 (l
1-C
"//
s2
S2
(s 2
(s 2
2CwB j
+
+
<o 2
where
7S)(i 2
(oa n
tan-
e-'/r)
vT"-
2f
$)
<o n2 2 )
ou2"
+ (On 2 )
(1
where
+
^
7" 2 a> 2
;(C0S 0>i/
COS CO2t)
<i
e-'/r
)
tan -1
<bT
=_J___
7
VI
+
rw
cos
(co n ,
" /r
^2
COS cot
eo 2 ) 2
<t>)
^r +
rcc->B
T)(l
-w sin
0) 2
2
2
C0l ){s
_ e
(a
tan
(1
7i)
e~
[aawT^TW - &)] __
-tan-.Z^TI^
2t
-C
1+aW
^L=
tar.-'
IT
(mVl
sin
flC(B
tan -1 acu
Vd-{C
x
as)
2r.cos
TcQnHa
a 2 co 2 e^"-'
?^lz_C! _
-
tan-i
a> 2 )
<j>
WJgl
tan-.
+
vi + r
-1 a>T
a< tan
rw
Vi^T
where
s(l
,,.,
-r/T
a?)
Ts)(s 2
(QH1
s(s 2
-C 2 \l
as)
2Z<os
Q> 2 (1
(1
1 1
(Qn
'l
0)
App. B
Laplace Transform
77me Funtion
F(s)
/CO
+ as + bs 2
+ ri5)(l +T2 5)
r,
T2 ) +
s 2 (l
ri
-f
Tl2 e-"r
b-aT2 + T2 2
r,
+ bs 2
s(s 2 .+ 2Ccos + m 2 )
co2(l
as
/(l-aCco-6co2+26C 2 co 2 ) 2
s2
C02)2
tan-
e -cnf sin
where
{s 2
o -
(av/l
#,-- ^1",
+ - aCco
Ao>(2C 2
1)
=^-(sin tat
,,r,
- r2
tan-
2/
^-C"
#)
<*
APPENDIX
function /(xi, x 2
is
it
g(x u x 2
x)
(constant).
g
.
x can be
f=f(xi, x 2
x)
Xg(x u x 2 ,...,
x)
(C-l)
where
= Lagrange's multiplier
there may be m constraint equations of the form
&(x) = gt
i=\,2,...,m
X
In general,
for
m<
n.
F =/(x) +
where X u X 2
(C-2)
the vector
[x u
in
A^,(x)
xz
X 2 g 2 (x)
x\
(C-3)
...
X mgm (x)
(C-4)
is
single-constraint case
2.
650
Form
the function
Maximize or
F = f(x) + Xg(x)
minimize F with respect to
(C-5)
x u x2
x; that
is,
App. C
651
simultaneously set
|^ =
i=l,2,...,
(C-6)
ox,
x,
3.
2,
i=\,2,...,n
/*,(A)
n, as
(C-7)
g(xi,x 2
Equation (C-8)
now gives
x)
(C-8)
x2
=
.
n.
2,
x obtained
from
The
justification
is
as follows. Let a
Then
maximum, minimum,
or saddle
x2
x) is
that
$fox,
(C-9)
x ,
constraint that
g(x u x 2 )
(C-10)
g (constant)
#=0
and
axi
=
J^
ax
(C-ll)
Let us assume that x^ and x 2 are arbitrary functions of a dependent variis, /and g can be written as/(x,, x 2 y) and g(x u x 2
Then, we can form the following equations
able y; that
df
dy
dg
dy
_^idxJ
dxi dy
dx 2 dy
dj^dx^ .dg^dxj
dx 2 dy
y), respectively.
(C-12)
(C-13)
dxi dy
,^Ldx 2
at the stationary
652
App. C
If
now X
is
so chosen that
& + *&-
(C- 17)
+*-
<
c- 18 >
and
g(x ,x 2 )
i
(C-19)
is satisfied.
Thus
Xg
F = f+
Index
Absolute
stability,
316, 365
multivariable systems, 62
AC
Actuating signal,
1,
Analytic function, 17
Asymptotic
stability,
317, 365
Cause-and-effect, 7
Back emf
286
Backlash, 200
Bandwidth, 463
ratio curve,
Complex
Complex
plane, 15
variable, 15
functions of, 16
loci,
277
653
654
Index
loci,
277
Constant
loci,
485
cascade, 139
277
137
direct,
N loci, 489
Constant
parallel,
Control system:
139
ac, 12
classification, 11
closed-loop, 3
continuous-data, 12
dc, 12
Digital controller, 14
digital, 13
open-loop, 2
sampled-data, 13
Controllability, 109, 144, 585,
587
z- transform solution,
165
output, 151
state,
447
Discrete state equation, 163
145
Controlled process,
decomposition, 169
Controlled variable,
Disturbance, 9
plot)
Coulomb
damping, 275
132
Cutoff rate, 463
Critical
CSMP,
Encircled, 333
Enclosed, 334
Damping
constant (see
Damping
Damping
Damping
state,
317
276
275
factor,
ratio,
Energy, 196
Equilibrium
factor)
210
220
permanent-magnet, 210
torque-speed curves, 225
Feedback:
effect
on external disturbance, 9
Index / 65S
Feedback (contd.)
effect on overall gain, 7
effect on sensitivity, 8
effect on stability, 8
effects of,
Initial- value
theorem, 21
Input node, 67
Input vector, 54, 99
Integral control,
300
574
574
basic elements, 5
block diagram, 3, 5
Inverse z-transform, 43
inversion formula, 44
types, 11
FORTRAN,
132
Friction, 191
Coulomb, 192
rotational motion, 195
static,
192
viscous, 192
equations, 25
definition, 18
inverse, 19
table,
Error series)
646
theorems, 19
Laplace transform
table,
646
equations, 158
604
infinite-time,
607
Linear systems, 11
Lyapunov's
stability criterion,
322
193
M
Magnification curves, 465, 468, 469,
472
6S6
Index
643
Mass, 190
Natural
Matrix:
32
addition,
adjoint of, 3
Nonlinear
32
algebra,
Nyquist
associate law, 33
state equation,
criterion, 322,
158
330
column, 28
commutative law, 33
conjugate, 31
definiteness, 38
definition,
27
determinant of, 29
diagonal, 28
35
inverse of, 35
multiplication, 33
nonsingular, 29
null,
29
Output
controllability, 151
order of, 28
Output equation, 97
quadrative form, 37
Output node, 67
rank of, 36
row, 28
singular,
29
skew- symmetric, 30
square, 29
subtraction, 33
symmetric, 29
transpose of, 30
unity, 28
Maximum
Mechanical
Mechanical
Mechanical
Mechanical
power
(see
Power)
356
Multivariable system, 2, 53
43
615
586, 617
Index
Root
380
construction,
Popov's
rules of construction,
363
303
Positional error constant, 264
447
411
528
Potentiometer, 209
Power, 196
Principle of argument, 335
Principle of optimality, 601
657
loci (contd.)
Poles, definition, 17
stability criterion,
Quadratic forms, 37
Sample-and-hold, 162
systems)
Ramp
input, 261
Sampler:
RC
RC
40
40
Sampling duration, 39
Sampling period, 39
finite-pulse width,
ideal,
Sensitivity, 7,
Sensitivity function, 9,
Servoamplifier, 246
algebraic,
499
284
algebra, 69
610
604
basic properties, 66
differential,
366
497
610
construction of, 71
definitions,
gain formula, 75
67
loop, 69
path,
68
Signal-to-noise ratio, 9
Similarity transformation, 118, 148
658
Index
316
Stability, 8,
316
absolute,
asymptotic, 317
definition,
317
316
587
Time delay
State concept, 95
State controllability, 112,
State diagram, 126,
145-146
274
Time-varying
460
state equation,
173
12, 173
188
matrix representation, 99
feedback)
State observer,
259
solution, 175
19
electrical networks,
analysis,
Time-invariant system, 12
canonical form,
Time-domain
588
Transfer function, 51
decomposition
133,317
of,
136
discrete-data systems, 81
linear systems, 51
317
properties, 102
96
99
friction, 192
State vector,
Static
coefficient, 192
211
Suppressed-carrier modulation, 212
Synchro, 213
Vandermonde
Index / 659
Viscous
friction,
192
Sample-and-hold)
zeros, definition, 17
z-transfer function, 83
z-transform, 15, 39
Weighting function, 55
Weighting sequence, 82
definition, 41
final-value theorem,
46
46
initial-value theorem,
43
inverse,
table,
43
theorems, 45
zero-input response, 317
KUO
PRLN iud
HALL
I
0-13-054973-8