This action might not be possible to undo. Are you sure you want to continue?
introduction to control systems 3
 feedback in control systems (click on any line to read the contents)19
Laplace transforms 41
Transfer Functions and its properties 80
Mason's Gain Formula 89
 Time Domain Analysis and steady state errors 113
 Concept of BIBO and Routh Harwitz criterion for stability 127
State variable Analysis 156
An introduction to controllers 169
 Root locus technique 193
Frequency Domain Analysis(polar and Bode plot) 206
Digital Control Systems 238
Z  Transforms 261
1
D.C. MOTOR SPEED CONTROL 272
Compensation 327
Control system in Biotechnology 350
2
INTRODUCTION
TO
CONTROL
SYSTEMS
BY
V DINESH BHARGAV ( 07359 )
MECHANICAL ENGINEERING
The following presentation consists of :
×
What is control system?
×
History of the control systems,
×
Introductory concepts and terms,
×
Types of control systems,
×
Comparison between open loop and
closed loop control systems,
×
Transfer function and its properties,
×
Types of signals and
×
Examples.
×
WHAT IS CONTROL SYSTEM?
Control system is an amalgamation or
combination of different physical components related
or connected together in such a manner to regulate,
direct , control itself to obtain desired output.
In the following slides the terms used in
control system are explained.
×
HISTORY OF CONTROL SYSTEMS:
Historically, the wide practical applications of control
systems first took place in the process areas like
petroleum processing, etc. Most were developed by the
successful realization by the intuitive and experimentally
oriented engineering methods.
The first known control system was the centrifugal
governor used for speed control of steam engine invented
by James Watt.
Later, people like Minorsky, Hazen and Nyquist
contributed to the development of the control systems.
In the year 1922, Minorsky invented automatic
controller for the ships.
In the year 1932, Nyquist developed the criterion for
stability of the closed loop systems.
In the year 1934, Hazen developed the theory of servo
mechanism, which is a comprehensive mathematical theory
aimed at applications where in there is need to control
mechanical motions viz., displacement, velocity, acceleration.
Around the second world war, the technical needs of military
like automatic airplane pilots , radar antennae etc, led to the
scientific approach in this field.
In the year 1940, Bode developed bode plot which made
linear closed loop control systems that satisfied performance
criteria possible.
In the year 1950, Evans developed root locus. Between 1960
to 1980, the control of deterministic and stochastic control
systems were achieved.
By the year 1980, research was being done modern control
systems and digital control systems. From the year 2000,
developments were made in the field of micro and nano
control systems.
WATT’S CENTRIFUGAL GOVERNOR
×
INTRODUCTORY CONCEPTS
AND TERMS:
System:
It is an arrangement or combination of different physical
components that are connected or related together to
form an entire unit to achieve a certain objective.
Control:
It is the process of regulation, direction and tracking.
Plant:
It is the portion of the system which needs to be
controlled, also called as controlled system.
Process:
It is a natural or artificial, voluntary progressing,
continuous operation or development marked by gradual
changes that succeed one another in relatively fixed way
to obtain a desired result.
Input:
It is a signal or excitation that is applied to a control system to get a
specified result. It is of two types:
Manipulated input: Input which is subject to our control.
Disturbance input: These are undesirable effects beyond our control
generated from surrounding environment.
Output:
It is the actual response from the control system due to the
application of specified input .
Disturbance:
It is any signal that has an adverse impact on the system.
Controller:
It is the element of the system itself or an external agency which
control the plant or process.
The words controller, regulator, tracking system are used in
different situations.
o
Controller:
The primary role of the controller is the clever
management of manipulated variables so as to
counteract the effects of disturbance for the specified
output.
o
Regulator:
When the desired value of the controlled outputs is
more or less fixed and the main problem is to reject
disturbance effects, the control system is called a
regulator.
o
Tracking system:
In the follow up or tracking system , the controlled
output is required to follow or track a time varying
command input.
Command input
Controlle
r
Process
Manipulated variable
Controlled output
Disturbance
Block diagram representing various parts of control systems
×
TYPES OF CONTROL SYSTEMS:
Based on various parameters, control systems are
classified in different ways as follows:
1. Open loop and closed loop control systems:
The systems in which output doesn’t have any impact on
the system are called open loop control systems.
The systems which maintains a prescribed relationship
between the output and the reference input by comparing
and using the difference as a means of control using
feedback is closed loop control system.
Feedback:
The arrangement of a system by which output is compared
with a input to reduce error to make system more stable is
called feedback.
Controlled
output
Command
input
Controller Process
Manipulated
variable
Disturbance
Open loop control systems:
Command
input
Controller Process
Manipulated
variable
Controlled
output
Disturbance
Measurement
Feedback signal
Closed loop control systems:
2. Stochastic and deterministic control systems:
If the output is known from input or external
disturbances it is known as deterministic control
systems.
If the output is unpredictable it is known as stochastic
control system.
3. Linear and non linear control system:
Linear control system is the one which obeys the
principle of superposition and homogeneity.
i.e., f( x+y ) = f( x ) + f( y ) ; f( kx ) = k f( x )
Non linear control system is the one which does not
obey the above principle. Linearization over desired
ranges of operation is possible.
4. Time varying and time invariant control systems:
When the parameters of the model vary with time it is
called time varying control system. Eg:– mass changes
should be considered in guidance of rocket systems.
If the parameters of the model doesn’t vary with time ,
they are called time invariant control systems.
5. Distributed and lumped parameter models:
In a distributed parameter models, the variables in a system
are distributed in space and vary with both the spatial
coordinates and time. These models consists of partial
differential equations w.r.t time and space.
In lumped parameter models, the matter is assumed to be
lumped at some discrete points of space. These models use
differential equations w.r.t time.
6. SISO and MIMO systems:
If the system has only single input and single output , it is
called single input single output system.
If the system has multiple inputs and outputs, it is called
multiple input and multiple output system.
7. According to application:
the control systems may be electrical, mechanical, thermal,
pneumatic, business, managerial control systems.
×
COMPARISON BETWEEN OPEN
LOOP AND CLOSED LOOP CONTROL
SYSTEM:
Depending on the application of control system, one may
either use open loop control system or closed loop control
system. The selection of the control system is also based upon
sensitivity criteria and economic considerations.
In the closed loop control systems , the effects of
feedback depend upon the kind of feedback used . There are
two types namely , positive feedback and negative feedback
Open loop control
system
Closed loop control
system
It is sensitive to external disturbances and
internal variations.
It is insensitive to external disturbances
and internal variations.
To get accurate results ,we need to use
highly expensive components.
To get accurate results ,we can use
inexpensive components also.
It is used when inputs are known and
there are minimal disturbances.
It is used where high sensitivity is required.
Calibration is easy . Recalibration is
needed from time to time.
Calibration is difficult. Generally
recalibration is not needed.
Generally very easy to build. Design of the system is much more
complicated.
Generally ,the problem of oscillations is
not witnessed.
To achieve accurate results , it may tend to
overcorrect errors leading to oscillations of
the system.
High in manufacturing , maintainance cost
and power consumption.
Low in manufacturing , maintainance cost
and power consumption.
Stability until unless disturbance is given. Wider range of stability.
The main differences between open loop and closed control
systems are:
FEEDBACK IN CONTROL SYSTEMS
By
V.V.Kaushik Sridhar,
3/4 B.Tech, MME,
Roll no. 7532
O
Based on whether feedback is given to a system or
not, control systems can be classified into:
1. Open control system (in which no feedback is
given).
2. Closed control system (in which feedback is given)
Both the above mentioned control systems have
their own advantages and disadvantages.
WHAT IS FEEDBACK?
O
The arrangement of a system by which output is
compared with the input to make the system
more stable or errorfree is known as feedback.
O
Feedback can also be defined as an arrangement
in which a part of the output is given back or fed
back into the input.
O
Based on the definition of feedback given in
the previous slide, one can define open loop
and closed loop control systems as given
below:
O
An open loop system may be defined as one
that is activated by a single signal i.e. the
input and the output has no effect on the
system.
O
A closed loop system can be defined as one
that is driven by two signals one the input
signal and the other, a signal called the
feedback derived from the output of the
system.
BLOCK DIAGRAM OF OPEN AND
CLOSED LOOP CONTROL SYSTEMS
Closed Loop Control System
TYPES OF FEEDBACK
O
There are two types of feedback namely positive
feedback and negative feedback.
O
Negative feedback occurs when the output of a
system acts to oppose changes to the input of the
system; with the result that the changes are
attenuated. If the overall feedback of the system is
negative, then the system will tend to be stable.
O
Positive feedback, sometimes referred to as
"cumulative causation", is a feedback loop system
in which the system responds to perturbation in
the same direction as the perturbation.
COMPARISON BETWEEN POSITIVE
AND NEGATIVE FEEDBACK
SYSTEMS
Positive feedback
1.Very rarely used.
2.Transfer function is given
by:
T.F.= G(s)
1G(s)H(s)
3. In this feedback, error
cannot be reduced but
overall gain of the
system increases.
Negative feedback
1.Very widely used.
2.Transfer function is given
by:
T.F.= G(s)
1+G(s)H(s)
3. In this feedback, error
can be reduced but
overall gain of the
system decreases.
EFFECTS OF FEEDBACK
O
Gain is reduced in –ve feedback and increased in
+ve feedback.
O
Parameter variation is reduced.
O
Sensitivity is increased.
The above effects of feedback help us to list
down the various advantages and disadvantages
of open loop and closed loop control systems.
ADVANTAGES OF OPEN LOOP AND
CLOSED LOOP CS
Open loop CS
1. Easy to design.
2. Cost factor more
economical.
3. More stable (no
oscillations till
disturbances).
4. No error signal.
5. Easy to maintain.
Closed loop CS
1. More accurate.
2. Can accomodate any
disturbance.
3. Wider stability.
DISADVANTAGES OF OPEN LOOP
AND CLOSED LOOP CS
Open loop CS
1. Cannot accommodate
disturbances.
2. Poor performance
under disturbances.
3. Re calibration
needed from time to
time.
Examples: Driving a
scooter, immersion
heater
Closed loop CS
1. Complex design.
2. Cost factorCostly.
3. Difficult to calibrate.
4. Oscillations.
Examples: Iron box, Air
conditioner.
EXAMPLES AND APPLICATIONS OF
FEEDBACK CONTROL SYSTEMS
O
Open loop control systems(nonfeedback):
Most automatic toasters are open loop control
systems because they are controlled by a timer.
This time required to make the toast should be
estimated by the user, who is not a part of the
system. Control over the quality of the toast (the
output) is removed once the time, which is bot
the input and the control action , has been set.
The time is typically set by calibrated means or a
switch.
O
Closed loop control systems:
An autopilot mechanism and the airplane it
controls is an example of a closed loop control
system. Its purpose is to maintain a specified airplane
heading, despite atmospheric changes. It performs
this task by continuously measuring the airplane
heading, and automatically adjusting the airplane
control surfaces ( rudder, ailerons) so as to bring the
actual airplane into correspondence with the specified
heading. The human pilot or operator who presets the
autopilot is not a part of the control system.
O
B
L
O
C
K
D
I
A
G
R
A
M
O
F
A
N
A
U
T
O
P
I
L
O
T
S
Y
S
T
E
M
T
H
A
T
U
S
E
S
F
E
E
D
B
A
C
K
C
O
N
T
R
O
L
O
Positive feedback control systems:
Audio feedback or acoustic feedback is
a common example of positive feedback. It
is the familiar squeal that results when
sound from loudspeakers enters a closely
placed microphone and gets amplified, and
as a result the sound gets louder and louder.
To avoid this condition, the microphone
must be prevented from "hearing" its
own loudspeaker.
O
Example of a biological positive feedback loop is the
process of blood clotting. The loop is initiated when
injured tissue releases signal chemicals that activate
platelets in the blood. An activated platelet releases
chemicals to activate more platelets, causing a rapid
cascade and the formation of a blood clot.
O
Negative feedback control systems:
1. A negative feedback amplifier, or more commonly simply
a feedback amplifier, is an amplifier which uses negative feedback to
improve performance (gain stability, linearity, frequency
response, step response) and reduce sensitivity to parameter
variations due to manufacturing or environmental uncertainties.
2. Its operation is most easily seen in its use by James Watt to control
the speed of his steam engine. Two heavy balls on an upright frame
rotate at the same speed as the engine. As their speed increases they
move outwards due to the centrifugal force. This causes them to lift a
mechanism which closes the steam inlet valve and the engine slows.
When the speed of the engine falls too far, the balls will move in the
opposite direction and open the steam valve.
3. Many biological process (e.g., in the human
anatomy) use negative feedback. Examples of
this are numerous, from the regulating of body
temperature, to the regulating of
blood glucose levels. The disruption of
feedback loops can lead to undesirable results:
in the case of blood glucose levels, if negative
feedback fails, the glucose levels in the blood
may begin to rise dramatically, thus resulting
in diabetes.
F
E
E
D
B
A
C
K
L
O
O
P
I
N
A
B
I
O
L
O
G
I
C
A
L
C
O
N
T
R
O
L
S
Y
S
T
E
M
×
EXAMPLES:
There are many examples for both open loop and closed loop control systems both in nature
and which are man made .
+
A MAN DRIVING A CAR:
Consider the possibility in which a man drives a car blind folded . Then he is safe as
long as there is no traffic on the road . But without his eyes while driving a car in traffic , he
can’t sense the mechanical motions like distance , speed , acceleration of his own vehicle and
that of others also . But when he is able to see , he can sense these motions . This information
is sent to brain which commands the limbs to act accordingly .
In the example discussed above , brain acts like a controller.
The motions which are sensed are the feedback signals . Therefore the case in which the man
drives with his eyes closed is an example for open loop control system . And the case in which
the man drives normally is an example for closed loop control system .
+
RESIDENTIAL HEATING SYSTEM:
Consider the case of the residential heating system , which can be
controlled by using both open loop and closed loop control system .
Consider the heating which makes use of open loop control system
, whose basic structure is shown below;
Temperature
set point
Valve
Room
radiator
Steam
Outdoor
temperature
Open loop model:
Controlled room
temperature
In this system , the indoor temperature is variable which is to be controlled
and the outdoor temperature is main disturbance input.
The desired is set on a calibrated dial , which positions the valve
that admits the steam for circulation through the radiator . The valve dial
is calibrated at a certain temperature . When the value changes
significantly the controlled temperature will deviate from the desired
value by a large error and hence precise control isn’t realized . The open
loop model is used for residential areas where control isn’t very stringent
The open loop model can be converted to a closed loop model by adding the function of
measurement of the controlled temperature. Comparison of measured and desired values of
temperature makes the control logic element to make changes in the heating rate so as to obtain
the desired value .
To control the above system , a thermostat with a bimetallic temperature sensor and a
snap action switch to control a simple solenoid actuated steam valve is often used .This
controls the power supply to the solenoid . Generally , two cases arise when the measured
temperature isn’t equal to the desired value .
If T , Ѳ represent desired and measured temperatures respectively , then
•When T < Ѳ i.e., T –Ѳ is positive then switch closes , solenoid energized , steam valve opens
and temperature rises .
•When T > Ѳ i.e., T –Ѳ is negative then switch opens , solenoid deenergized , temperature
falls down due to heat transfer to surroundings .
LAPLACE TRANSFORMS
V.Vasudeven
07360
Mechanical department
WHAT ARE LAPLACE TRANSFORMS?
A Laplace transform is a type of
integral transform.
Plug one function in
0
s t
e dt
∞
− ⋅
∫
( ) f t
Get another function out
( ) F s ·
The new function is in a different domain.
( ) F s
is the Laplace transform of
( ). f t
Write
{ ¦
( ) ( ), f t F s · L
0
s t
e dt
∞
− ⋅
∫
( ) f t ( ) F s ·
When
{ ¦
{ ¦
( ) ( ),
( ) ( ), etc.
y t Y s
x t X s
·
·
L
L
A Laplace transform is an example of an improper integral : one of
its limits is infinite.
0 0
( ) lim ( )
h
s t s t
h
e f t dt e f t dt
∞
− ⋅ − ⋅
→∞
·
∫ ∫
Define
A CALCULATION
Let
0 if
( )
1 if
t c
u t c
t c
<
¹
− ·
'
≥
¹
This is called the unit step function or
the Heaviside function.
It’s handy for describing functions that
turn on and off.
c
1
t
0 if
( )
1 if
t c
u t c
t c
<
¹
− ·
'
≥
¹
The Heaviside Function
{ ¦
0
1 1
( ) ( ) lim
lim lim ( )
h
s t s t
h
c
h s c
s t s h s c
s s
c
h h
u t c e u t c dt e dt
e
e e e
s
∞
− ⋅ − ⋅
→∞
− ⋅
− ⋅ − ⋅ − ⋅
− −
→∞ →∞
− · − · ·
· − ·
∫ ∫
L
Calculating the Laplace transform of the
Heaviside function is almost trivial.
Remember that
( ) u t c −
is zero until
then it’s one.
, t c ·
We can use Laplace transforms to turn an
initial value problem
" 3 ' 4 ( 1)
(0) 1, '(0) 2
y y y t u t
y y
+ − · ⋅ −
·− ·
into an algebraic problem
2
2
1
( )*( 3 4) ( 1)
s
s
s e
Y s s s s
+
⋅
+ − + + ·
Solve for y(t)
Solve for Y(s)
1
1
A sawtooth function
t
Laplace transforms are particularly effective
on differential equations with forcing functions
that are piecewise, like the Heaviside function,
and other functions that turn on and off.
I.V.P.
Laplace transform
Algebraic Eqn
If we solve the algebraic equation
2
2 2
( 1) ( 1)
( )
( 3 4)
s s
s s e e
Y s
s s s
−
− + ⋅ ⋅ − ⋅
·
⋅ + −
and find the inverse Laplace transform of the solution, Y(s), we have the
solution to the I.V.P.
Algebraic Expression
Soln. to IVP
Inverse
Laplace
transform
The inverse Laplace transform of
is
4
4
3 3 2 1
5 80 4 16
4
3 2
5 5
( ) ( 1)( + ( ) )
( )( ( ) )
t t
e
e
t t
y t u t e e t
u t e e
−
−
· − ⋅ ⋅ − −
− ⋅ − ⋅
2
2 2
( 1) ( 1)
( )
( 3 4)
s s
s s e e
Y s
s s s
−
− + ⋅ ⋅ − ⋅
·
⋅ + −
4
4
3 3 2 1
5 80 4 16
4
3 2
5 5
( ) ( 1)( + ( ) )
( )( ( ) )
t t
e
e
t t
y t u t e e t
u t e e
−
−
· − ⋅ ⋅ − −
− ⋅ − ⋅
is the solution to the I.V.P.
" 3 ' 4 ( 1)
(0) 1, '(0) 2
y y y t u t
y y
+ − · ⋅ −
·− ·
Thus
HOW DO WE TRANSFORM AN
DIFFERENTIAL EQUATION?
we need several nice properties of Laplace
transforms that may not be readily apparent.
First, Laplace transforms, and inverse
transforms, are linear :
{ ¦ { ¦ { ¦
{ ¦ { ¦ { ¦
1 1 1
( ) ( ) ( ) ( ) ,
( ) ( ) ( ) ( )
cf t g t c f t g t
cF s G s c F s G s
− −
+
+
L = L +L
L = L +L
for functions f(t), g(t), constant c, and
transforms F(s), G(s).
there is a very simple relationship
between the Laplace transform of a given
function and the Laplace transform of that
function’s derivative.
{ ¦ { ¦
{ ¦ { ¦
2
'( ) ( ) (0),
''( ) ( ) (0) '(0)
f t s f t f
f t s f t s f f
⋅ −
⋅ − ⋅ −
L = L
L = L
and
These show when we apply differentiation
by parts to the integral defining the transform.
Second,
DIFFERENTIAL EQUATIONS
1 0
) 0 ( ' ) 0 ( ) ( ' " K y K y t r by ay y · · · + +
) ( )] 0 ( [ )] 0 ( ' ) 0 ( [
2
s R bY y sY a y sy Y s · + − + − −
) ( ) ( ) ( )] 0 ( ' ) 0 ( ) [( ) (
1
) (
2
s Q s R s Q y y a s s Y
b as s
s Q
+ + + ·
+ +
·
) ( ) (
1
Y L t y
−
·
1
st
step
2
nd
step
3
rd
step
Now we know there are rules that let
us determine the Laplace transform
of an initial value problem, but...
HOW DO WE FIND INVERSE
LAPLACE TRANSFORMS?
First you must know that Laplace transforms
are onetoone on continuous functions.
In symbols
{ ¦ { ¦
( ) ( ) ( ) ( ) f t g t f t g t ⇒ · L =L
when f and g are continuous.
That means that Laplace transforms are
invertible.
INVERSE LAPLACE TRANSFORMS
If
{ ¦
( ) ( ), f t F s · L
{ ¦
1
1
2
( ) ( )
c i
s t
i
c i
F s e F s ds
π
+∞
− ⋅
−∞
·
∫
L
then
{ ¦
1
( ) ( ), F s f t · L
where
An inverse Laplace transform is an improper
contour integral, a creature from the world
of complex variables.
That’s why you don’t see them naked very often. You usually just see what they
yield, the output.
In practice, Laplace transforms and inverse
Laplace transforms are obtained using tables
and computer algebra systems.
IMPULSE?
An impulse is the effect of a force that acts over a very short time interval.
Engineers and physicists use the Dirac delta function to model impulses.
A lightning strike creates an electrical
impulse.
The force of a major leaguer’s bat
striking a baseball creates a mechanical
impulse.
THE DIRAC DELTA FUNCTION
This socalled quasifunction was created
by P.A.M. Dirac, the inventor of quantum
mechanics.
0
( ) 0 ( ) 1 t a t a t a dt δ δ
∞
− · ≠ − ·
∫
when and
People use this thing all the time. You
need to be familiar with it.
THE LAPLACE TRANSFORM OF THE
DIRAC DELTA FUNCTION
{ ( )} 1/
a s
L t a e δ
⋅
− ·
DIRAC DELTA (UNIT IMPULSE) FUNCTION
O
Impulse of f(t), for
O
“Generalized function”
O
Laplace transform
∫
∞
· −
¹
'
¹
· ∞
· −
0
1 ) ( and
otherwise 0
if
) ( a t
a t
a t δ δ
as
e a t L
−
· − )} ( {δ
1 0
t t t ≤ ≤
t
t
Laplace transforms have limited appeal.
You cannot use them to find general solutions
to differential equations.
You cannot use them on initial value problems
with initial conditions different from
1 2
(0) , '(0) , y c y c · · etc.
Initial conditions at a point other than zero
will not do.
LAPLACE TRANSFORMS
O
Def:
O
Inverse:
O
Linearity:
O
Shifting Theorom:
0 ), ( for ) ( ) ( ) (
0
> · ·
∫
∞
−
t t f dt t f e f L s F
st
) ( ) (
1
F L t f
−
·
)} ( { )} ( { )} ( ) ( { t g bL t f aL t bg t af L + · +
)} ( { ) (
) ( )} ( {
1
a s F L t f e
a s F t f e L
at
at
− ·
− ·
−
REVIEW: PARTIAL FRACTIONS
n
n k
k k
n
k
m
m
m
p s
A
p s
A
p s
A
p s
A
p s
A
p s p s p s
a s a s
−
+
−
+
−
+
−
+
−
·
− − −
+ +
−
−
−
· ·
·
·
) ( ) ( ) ( ) (
) ( ) ( ) (
2
2
1
1
1
1
12
1
11
2 1
0
1
1
•
Case I: unrepeated real factor
•
Case II: repeated real factor
•
Case III: unrepeated complex factor
•
Case IV: repeated complex factor
n
n k k
k k
n
k
m
m
m
p s
A
p s
A
c bs s
B s A
c bs s
B s A
c bs s
B s A
p s p s c bs s
a s a s
−
+
−
+
+ +
+
+
+ +
+
+
+ +
+
·
− − + +
+ +
−
−
−
· ·
·
·
) ( ) ( ) ( ) (
) ( ) ( ) (
2
2
2
1 1
1 2
12 12
2
11 11
2
2
0
1
1
EXISTENCE FOR LAPLACE TRANSFORM
•
piecewise continuous on every finite inteval in the
range of
•
satisfy
for some constant k and M.
Conclusion:
•
The Laplace transform of exists for all s > k
0 all for  ) (  ≥ ≤
−
t Me t f
kt
) (t f
) (t f
0 ≥ t
) (t f
Condition:
TRANSFORM OF DERIVATIVES
O
THEORM 1
¤
Laplace of f(t) exists
¤
f’(t) exists and piecewise continuous for t>=0
O
THEOREM 2
) 0 ( ) ( ) ' ( f f sL f L − ·
) 0 ( ) 0 ( ' ) 0 ( ) ( ) (
) 1 ( 2 1 ) ( − − −
− − − − ·
n n n n n
f f s f s f L s f L ·
TRANSFORM OF INTEGRALS
s
s F
s
f L
d f L
t
) ( ) (
} ) ( {
0
· ·
∫
τ τ
}
) (
{ ) (
1
0
s
s F
L d f
t
−
·
∫
τ τ
CONVOLUTION
O
Definition
O
Property
O
Application
∫
− · ∗ ·
t
d t g f t g f t h
0
) ( ) ( ) )( ( ) ( τ τ τ
)} ( ) ( { ) )( * (
1
s G s F L t g f
−
·
) ( ) ( ) (
1
) (
2
s Q s R s Y
b as s
s Q
·
+ +
·
∫
− ·
t
d t t q t y
0
) ( ) ( ) ( τ τ τ
0 ) 0 ( ' 0 ) 0 ( ) ( ' " · · · + + y y t r by ay y
1. Solve
2. Calculate integral
List of Laplace Transforms
f(t) L(f) f(t) L(f)
1
1
1/s
7
cos ω t
2
t
1/s
2
8
sin ω t
3
t
2
2!/s
3
9
cosh at
4
t
n
(n=0, 1,…)
10
sinh at
5
t
a
(a positive)
11
e
at
cos ω t
6
e
at 12
e
at
sin ω t
1
!
+ n
s
n
1
) 1 (
+
+ Γ
a
s
a
a s −
1
2 2
ω + s
s
2 2
ω
ω
+ s
2 2
a s
s
−
2 2
a s
a
−
2 2
) ( ω + −
−
a s
a s
2 2
) ( ω
ω
+ −a s
LINEAR SYSTEM APPLICATIONS
O
Definition: A linear mapping of a set of input functions
into a corresponding set of output functions
O
Transfer function: ratio of the Laplace transform of the
output of the system to that of the input of the system
)} ( { ) ( t x f t y ·
TRANSFER FUNCTION PROPERTIES
O
Completely characterizes the systems (initially
relaxed)
O
Independent of inputs and outputs
O
Equals the Laplace transform of the impulse
response of system
O
Can be used to find the response to another system
input (using convolution or partial fraction)
THANK YOU
×
TRANSFER FUNCTION AND ITS
PROPERTIES:
Transfer function of a system is a mathematical
model and an operational method of expressing
differential equations that relates the output variable
to the input variable. It is the property of the system
itself which is independent of the magnitude of the
input or the driving function.
Transfer function is the ratio of Laplace transform
of output to the Laplace transform of input with initial
conditions set to zero.
Transfer function= C(s)/R(s), initial conditions zero.
O
Poles and zeros of transfer function:
C(s)/R(s) = ( b
o
s
n
+ b
1
s
n1
+ _ _ _ + b
n
) / ( a
o
s
m
+ a
1
s
m1
+ _ _ _ + a
m
)
= N(s) / P(s).
The poles of the transfer function are the roots of
the equation P(s)=0. The equation P(s)=0 is called
the characteristic equation of the system.
The zeros of the transfer function are the roots of
the equation N(s)=0.
O
Order and type of the system:
C(s)/R(s) = ( b
o
s
n
+ b
1
s
n1
+ _ _ _ + b
n
) / ( a
o
s
m
+
a
1
s
m1
+ _ _ _ + a
m
)*s
r
= N(s) / P(s)
Order of the system is the degree of the equation
P(s)=0.
Type of the system is the number of poles at the
origin.
TYPES OF INPUT SIGNALS
submitted by Tushar Ranjan Moharana
Date 9
th
September 2009
AssignmentI
BASIC TYPES
Basically inputs are classified into four groups :
1. Step input
2. Ramp input
3. Parabolic input
4. Impulse input
STEP INPUT
Q
Magnitude of input is constant(let 0) for certain time and suddenly increases to
some other value (let 1) and maintained at that value.
Q
Use : The functioning of automobiles on smooth road is tested by using
this input through
Simulation
Q
Used to test signal for
Electronic equipment.
Q
By giving step input signal
We can determine the
response of system when
it is switched on or off.
X –axis (time)
RAMP INPUT
Q
Impulse varies linearly with time.
Q
It is used in the horizontal deflection plates of a cathode ray
oscilloscope.
Q
It moves the electron beam horizontally with a constant speed
as the vertical deflection
plate trace the wave
form
X –axis (time)
PARABOLIC
INPUT
Q
Magnitude of input signal is proportional to
square of the time
Q
Used in accelerating speed of vehicle etc.
X –axis (time)
IMPULSIVE
INPUT
Q
Impulse have high magnitude but acts for less
time
X –axis (time)
Q
It is used to test the
response of equipments to
sudden fluctuation in the
input
Q
By giving impulse input
to any system we can
determine its behavior to any
other signal
O
Parabolic signal:
f(t)=t
2
/2, when t > 0
=0, otherwise.
Ramp signal and parabolic signal are used for the analysis
of faster signals.
O
Sinusoidal signal:
f(t)=sin(ωt)*μ(t)
Parabolic Signal
Ramp Signal
Step Input Signal
Impulse Signal
Sinusoidal Signal
MASON’S GAIN
FORMULA
NAMESINGO SABATINI
HEMBRAM
BRANCHMETALLURGICAL
AND MATERIALS
ENGINEERING
ROLL NUMBER 7528
MASON’S GAIN
FORMULA
To reduce a block diagram, one may use the Mason’s gain formula
(also called Mason’s rule).
Mason first derived the idea using what he called a signalflow
graph, which is a different
graphical representation of a block diagram. A signalflow graph is
drawn with paths (lines) and
nodes. The transfer functions in a block diagram become the paths
and the variables in between
the blocks become the nodes. The input and output variables of a
block diagram are designated the
source and sink nodes
Signal flow graph is a diagram
which represents a set of simultaneous
linear algebraic equation When applying
the signal flow graph method to control
system we must first transform linear
differential equation into algebraic
equation in ‘s’.
Signal flow graph consist of a network, in
which nodes are connected by directed
branches.
Each node represents a system variable and
each branch connected between two nodes
acts as a signal multiplier.
Note that signal flows only in one direction.
The direction of the signal flow is indicated
by an arrow placed on the branch and the
multiplication factor is indicated along the
branch.
The signal flow graph depicts the flow of signals
from one point of a system to another and gives the relationship
among the signals.
Signal flow graph contains essentially the same information as a
block diagram. The advantage of using
signal flow graph is to represent a control system is that a gain
formula or Mason’s gain formula is available
which gives the relationships among the system variables
without requiring a reduction of the graph.
TERMINOLOGY
x1 x2 x3
x4
Node: A node is a point representing a
variable signals.
Input node or source :It is a node which has
only outgoing branches. This corresponds
to an independent variable .Here x1 is source
node.
Output node or Sink : An output node or sink
is a node which has only incoming branches.
This corresponds to a dependent variable.
Here x4 is sink node.
A B
C
•
The gain of the branch is a transmittance
•
Branch :It is a directed line segment
joining two nodes.
Transmittance :It is a gain between the two
nodes.
Feed back loop: If a loop originates and
terminates at the same node ,is known as
feed back loop.
Self loop: A loop that consists of only one
node where it originates and terminates.
TERMINOLOGY
Chain node: In this node no incoming or outgoing .Here x2 and x3
are chain nodes.
Flow of signals: Source to sink.
Feed back: Sink to sour.
•
Dummy nodes: If incoming and outgoing branches exists at the
first and last nodes representing input and output variables can not
be considered input and output nodes.
•
In determining forward and backward paths self loops are not
considered.
•
In such cases separate input and output nodes are created by
adjoining branches with unity gain.
•
Path : It is a traversal of
connected branches in the
direction of the branch arrows.
•
Path gain: The product of all
gains going to the forward
path.
•
Loop :It is a closed path .
•
Loop gain :It is the product of the
branch transmittance of a loop.
•
Non touching loops : Loops
are nontouching if they do not
posses any common nodes.
Non touching loop
Touching loop
x1
x
2
x3
x4
x2
x1 x3 x4
x5
Forward path : A
forward path is a path
from an input node to
an output node which
does not cross any
nodes more then once.
•
Forward path gain : It is the
product of the branch
transmittances of a forward
path.
RULES OF SIGNAL FLOW
GRAPH
ADDITION RULE:
x2 b x3
c x4
X6 e d
x5
x1=ax2+bx3+cx4+dx5+ex6
x1
TRANSMISSION RULE:
x2 x5=dx1
a x4=cx1
x5 d b x2=ax1
c x3=bx1
x4
MULTIPLICATION RULE:
a b c abc
x3
x1
x2 x
1
x1 x
2
x3 x4
SFG VS BLOCK DIAGRAMS
O
Source or
input node
O
Sink or output
node
O
Branch
O
Node
O
Input signal
O
Output signal
O
Block
O
Signal
Signal Flow Graph
Block Diagram
PROPERTIES OF SIGNAL
FLOW GRAPH
1. A branch indicates the functional dependence of one signal
upon another. A signal passes through only
in the direction specified by the arrow of the branch .
2. A node adds the signals of all incoming branches and
transmits this sum to all outgoing branches.
3. A mixed node which has both incoming and outgoing
branches may be treated as an output node by adding an
outgoing branch of unity transmittance.
4. For a given system, signal flow graph is not unique. Many
different signal flow graphs can be drawn for a given system by
writing the system equations differently.
Utility of Signal Flow
Graphs
• Alternative to block diagram approach
 may be better for complex systems.
 good for highly interwoven systems.
 system variables represented as nodes, branches(lines) between
nodes show relationships between system variables.
• The “flow graph gain formula” (Mason)allows the system transfer
function to be directly computed without manipulation or reduction
of the diagram.
We now need to define some
more terms and show how each
of these quantities can be
calculated:
System determinant:
Δ = 1 – (sum of all individual loop gains)
+ (sum of the products of the gains of all possible two loops that
do not touch each other)
– (sum of the products of the gains of all possible three loops
that do not touch each other)
+ … and so forth with sums of higher number of nontouching loop
gains.
EXAMPLE
Step 1: forward paths=2
G
6
G1 G
2
G4 G3
G
5
H1
H2

H3
G1G2G3G4G
5
G
6
1 1
P1=G1G2G3G4G5
P2=G6
Step 2: No. of loops=4
Sum of product gain of two non touching loops=G4G6H2H3 +
G2G6H1H3 + G2G4H1H2
Step 3: three non touching loops=1
=L1L2L4= G2G4G6H1H2H3
H1
G
2 L1
G
4
H2
L
2
G1G2G3G4G
5
L
3
H3
G
6
L
4
H4
PRODUCT GAIN:
Δ=1 – (  G2H1G4H2
G1G2G3G4G5H3G6H3)
+
(G2G6H1H3+G4G6H2H3+G2G4H1H2
)
 (G2G4G6H1H2H3)
Δ1=Product gain(excluding P1)=1
Δ2=Product gain(excludinP2)=1(
G2H1G4H2)+
(G2G4G6H1H2H3)
=1+G2H1+G4H2+G2G4H1H2
TRANSFER FUNCTION:
The ratio of laplace transform of
output to laplace transform of
input at zero initials.
T.F=C(S)/R(S)
at zero initials
C(S)/R(S)=((P1*Δ1)+
(P2*Δ2))/Δ
HOW TO DEVLOP SFG FOR THE SYSTEM REPRESENTED BY THESE
EQUATIONS .
5X12X23X3=0
2X1+X2+1.5X4=0
2X2+X33X1=0.5X3
X4=2X3
Arrange the above equations :
X1=52X23X3
X2=2X11.5X4
X3=0.5X3+3X1+2X2
X4=2X3
HERE ‘5’ IS INPUT.
X1 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
X4
2
5
1.5
0.5
2
3
X2 X
3
2
3
DISADVANTAGES
•
Missing a critical loop or
probability of missing a
critical loop ,reduces the
accuracy.
•
Does not give a idea of
cause and effect of the
overall system.
FLOW GRAPH SUMMARY
•
Block diagrams and signal flow graphs allow visual
representation of complex dynamic systems.
•
Through Mason’s gain formula, we can compute the input to
output transfer function of the overall system.
•
Note how easily other transfer functions can be obtained e.g.
noise response –they all share the same Δ.
TIME DOMAIN ANALYSIS
The variation of output/response of a system wrt
time is called time response. It has two parts
Steady state response and transient response.
TRANSIENT RESPONSE: THE
RESPONSE WHICH GOES TO ZERO
AFTER LARGE INTERVAL OF TIME IS
TRANSIENT RESPONSE.
From this response we can
observe
1.Time taken by the system
to respond once the input is
given to the system.
2.The presence of
oscillations in the output.
STEADY STATE RESPONSE:THE
RESPONSE WHICH REMAINS EVEN
AFTER THE TRANSIENTS HAVE DIED
OUT.
From this response we can observe
1.Time taken by the system to reach steady state value.
2.Whether there error is existing in the output or not
STEADY STATE ERROR IN TIME
DOMAIN ANALYSIS
Þ
One of the objectives of most control systems is that the
system output response follows a specific reference signal
accurately in the steady state.
Þ
Steadystate error is the difference between the output
and the reference in the steady state.
Þ
Steadystate errors in control systems are almost
unavoidable and generally derive from the imperfections,
frictions, and the natural composition of the system.
Þ
In the design problem, one of the objectives is to keep the
steadystate error below a certain tolerable value.
O
The error in control system is defined as
e(t) = r(t)  b(t)
or
E(s) = R(s)  B(s) =R(s)  H(s)C(s)
The steadystate error is then given by:
The steadystate error can be evaluated using the final
value theorem as:
This shows that the steadystate error depends on the
reference input R(s) and the loop transfer function
G(S)H(S)
The open loop transfer function, G(S)H(S), may be
written as
Steadystate error due to a step input Steadystate error due to a step input
For R(s)
G(s)H(s) as the positional error constant, then
We can summarize the steadystate error due to a step input as follows:
Type0 system
(or higher) Type1 system
Steadystate errors: (a) step
input, (b) ramp input
Steadystate error due to a ramp input
Steadystate error due to a ramp input
· K lim lim
v
K
;n=0,1,2,..
n1
s 0 s 0
s
sG(s)H(s)=
→ →
For R(S)=
, we have
If we define If we define = velocity error constant = velocity error constant
Therefore,
Hence, we have
Hence, the following conclusions can be made:
The steadystate error could be written as:
Defining the acceleration error constant as:
The steadystate error of the system is:
Laplace transform of r(t) is given by:
The input is described by:
Steadystate error due to a parabolic input Steadystate error due to a parabolic input
Steadystate errors for various inputs and system types are
tabulated below.
•
The error constants for nonunity feedback systems may be obtained by
replacing G(s) by G(s)H(s).
•
Systems of type higher than two are not employed for two reasons:
1. The system is difficult to stabilize.
2. The dynamic errors for such systems tend to be larger than those
types0, 1 and 2.
GENERALISED COEFFICIENT METHOD
By using convolution theorem and taylor’s series
expansion we can write steady state error as
Es=k0R(t)+k1dR(t)/dt+……..+
kn d^nR(t)/dts
Where kn= lt s 0 d^n(1/1+G(s)H(s))
ds^n
MERITS:
# THIS METHOD GIVES
IMPORTANT INFORMATION
ABOUT VARIATION OF ERROR
WITH TIME.
Demerits:
# It gives error of only
standard signals.
# the value of error at zero
or infinity is not given by
this method.
# Applicable only for stable
systems.
CONCEPT OF BIBO
By : Alok Ranjan
WHY INVESTIGATE STABILITY?
O
The issue of ensuring the stability of a closed
loop system is the most important to control
system design. An unstable feedback system is
of no practical value. A stable system should
exhibit a bounded output if the corresponding
input is bounded. This is known as BIBO stability.
CONCEPT OF BIBO
º
A system is defined to be BIBO stable if every
bounded input to the system results in a
bounded output over the time interval zero to
infinity.
º
BIBO stability is based on transfer function
representation.
º
For LinearTimeinvariant (LTI) system, if the
impulse response is absolutely integrable,the
system is BIBO stable.
º
Impulse response approach to zero when time
lead to infinite.
CRITERION FOR BIBO STABILITY FOR LTIC
SYSTEMS
x(t) y(t)
y(t) = h(t)*x(t)
= x(t )d
= ≤
d
h(t)
∙
If x(t) is bounded, i.e. < K< ,then
< K < , and ≤ K d
Hence for BIBO stability.
d <
Q
The concept of BIBO stability is not appropriate to
describe stability in all nonlinear systems.
LTIC SYSTEM
R(s) C(s)
T(s)
T(s) = P(s)/Q(s) , Q(s)=a
n
)
C(s)=T(s)R(s)=P(s)/(a
n
))*R(s) = K
1
/(sp
1
) +.........
+K
n
/(sp
n
) +C
r
(s)
Characteristic equation : Q(s)=0
System roots or system poles : Pi , i =1,…….,n
Forced response : Cr(s) (the sum of terms , in the
partial fraction expansion, that originate in the poles of
R(s))
Q
A linear time invariant system is BIBO stable provided
all roots of the system characteristic equation (poles of
the closedloop transfer function) lie in the left half of
the splane.
EXAMPLE
O
Consider the following system:
O
=x+u
O
y=x
Transfer function of this system is 1/(s1).
O
Pole is at s=1 i.e. system is BIBO unstable.
O
Solution of this system is
O
Y(t)=e
t
x(0) +e
t
) d
If u( ) =1 and x(0)=1, we get y(t)=1.In other words
y(t) is bounded all time.However,this does not mean
that the system is BIBO stable!
BIBO stability requires that the output remain bounded
for all time, for all initial conditions and input  not just
for some specific initial conditions and input.
If we put u( )=0and x(0)=1then y(t)= e
t
which
approaches infinity with time, which proves that
system is BIBO unstable.
RELATIONSHIP BETWEEN BIBO AND
ASYMPTOTIC STABILITY
The Asymptotic stability of system implies that
system is also BIBO stable, but not vice versa.
because poles of G(s) is subset of Eigen values
of system matrix A(say).
A marginally stable or Asymptotically unstable
system is BIBO unstable.
In the absence of pole zero cancellations,BIBO
stability and Asymptotic stability are equivalent.
Two stabilities differ when RHP polezero cancellations
occur (including imaginary axis polezero cancellation)
Example:
= x + u
y= x
Asymptotic stability:
Eigen values :1,2 unstable
BIBO stability
T(s)=
poles =2 stable
EXAMPLES
O
Stability and supermaneuverability of modern
fighter aircraft(figure1 )
O
Vibration control (sputnik 1 and 2)(figure2 and 3)
STABILITY IN MODERN FIGHTER
AIRCRAFT
SPUTNIK 1
SPUTNIK 2
ROUTH – HURWITZ STABILITY CRITERION:
This criterion is based on ordering the coefficients
of the
characteristic equation into an array , called the Routh
array as:
q(s)=a
0
s
n
+ a
1
s
n1
+ a
2
s
n2
+……..+ a
n1
s + a
n
= 0
s
n
: a
0
a
2
a
4
a
6
s
n1
:
s
n2
:
.
a
1
b
1
.
a
3
b
2
.
a
5
b
3
. . .
s
2
:
s
1
:
s
0
:
e
1
f
1
a
n
a
n
s
4
s
3
s
0
s
1
s
2
1 18
16 8
5
8*18 – 1*16
8
= 16
8*5 – 1*0
8
0
16*16 – 8*5
16
= 13.5
= 5
5
0
:
:
:
:
:
Ex:1) Consider the fourthorder system with characteristic equation
s
4
+ 8s
3
+ 18s
2
+ 16s + 5 = 0
The Routh array for this system is as below:
The elements of the first column are all positive and hence the
system is stable.
The coefficients b
1
, b
2
,… are evaluated as follows:
b
1
= (a
1
a
2
– a
0
a
3
)/ a
1
;
b
2
= (a
1
a
4
– a
0
a
5
)/ a
1
;……
this process is continued till we get a zero as the last coefficient
in the third row. In similar way, the coefficients of 4
th
, 5
th
,….n
th
and (n+1)
th
rows are evaluated, e.g.,
c
1
= (b
1
a
3
– a
1
b
2
)/ b
1
;
c
2
= (b
1
a
5
– a
1
b
3
)/ b
1
;…… and
d
1
= (c
1
b
2
– b
1
c
2
)/ c
1
;
d
2
= (c
1
b
3
– b
1
c
3
)/ c
1
;……
It is to be noted here that in the process of getting the Routh
array, the missing terms are regarded as zeros. Also all the
elements of any row can be divided by a positive constant during
the process to simplify the computational work.
ROUTHHURWITZ STABILITY CRITERION:
The RouthHurwitz stability criterion is stated as:
For a system to be stable, it is necessary and
sufficient that each term of first column of Routh array of its
characteristic equation be positive if a
0
> 0. If this condition is
not met, the system is unstable and number of sign changes of
the terms of the first column of the Routh array corresponds to
the number of roots of the characteristic equation in the right
half of the s – plane.
Ex:2) Consider the fourthorder system with characteristic equation
3s
4
+ 10s
3
+ 5s
2
+ 5s + 2 = 0
The Routh array for this system is as below:
s
4
s
3
s
0
s
1
s
2
3 5
5 10
2
10*5 – 3*5
10
= 3.5
10*2 – 3*0
10
3.5*5 – 2*10
3.5
= 0.5/3.5
= 2
2
:
:
:
:
:
By examining the elements of the first column, it is found that
there are two sign changes. Therefore the given system is unstable
having two poles in the right half of splane.
Special cases:
Occasionally, in applying the Routh stability criterion,
certain difficulties arise causing the breakdown of the Routh’s
test. The difficulties encountered are generally of the following
types:
Difficulty 1: When the first term in any row is zero while rest of
the row has at least one non zero term.
Because of this zero term, the terms in the next row
become infinity and Routh’s test breaks down. The following
method is used to overcome this difficulty:
O
Substitute a small positive number є for the zero and proceed to
evaluate the rest of the array. Then examine the signs of the first column
by limiting є→0.
Ex:3) Consider the following with characteristic equation
s
5
+ s
4
+ 2s
3
+ 2s
2
+ 3s + 5 = 0
The Routh array for this system is as below:
s
4
s
3
s
0
s
1
s
2
1
2
2 є
5
2*є +2
є
→ ∞ 5
4є4 5є
2
2є+2
→2
5
:
:
:
:
:
s
5
1
: 2 3
→ 0
It is seen that since the first element in the third row is zero it is
replaced by є and proceeded. There are two sign changes in the
first column. Therefore the given system is unstable having two
poles in the right half of splane.
Difficulty 2 : When all the elements in any row of the Routh array
are zero.
This condition indicates that there are symmetrically
located roots in the splane .The polynomial whose coefficients
are the elements of the row just above the row of zeros in the
Routh array is called an auxiliary polynomial. Because of a zero
row, the Routh’s test breaks down. This situation is overcome by
replacing the row of zeros in the Routh array by a row of
coefficients of the polynomial generated by taking the first
derivative of the auxiliary polynomial.
Ex:4) Consider the following with characteristic equation
s
6
+ 2s
5
+ 8s
4
+ 12s
3
+ 20s
2
+ 16s + 16 = 0
The Routh array for this system is as below:
s
5
s
4
s
3
1
12
0
2
0
12
16
:
:
:
s
6
2
: 8 20 16
16
Since the terms in the third row are zero the Routh’s test breaks.
Now the auxiliary polynomial is formed from the coefficients of
the s
4
row, which is given by,
A(s) = 2s
4
+ 12s
2
+ 16
The derivative of the polynomial with respect to s is
d
ds
A(s) = 8s
3
+ 24s
The zeros in the third row are now replaced by the coefficients 8 and 24. The Routh
array then becomes
s
5
s
4
s
3
1
12
24
2
8
12
16
:
:
:
s
6
2
: 8 20 16
16
s
1
s
0
6
16
:
:
s
2
8/3
: 16
It is seen that there is no sign change in the first column of the new array. By
solving for the roots of auxiliary polynomial
2s
4
+ 12s
2
+ 16 =0
We find that the roots are s=±j√2 and s=±j2. These two pairs of roots are also
the roots of original characteristic equation. Since there is no sign in the new
array formed with the help of the auxiliary polynomial , we conclude that no
root of characteristic equation has positive real part. Therefore the system under
consideration is limitedly stable.
APPLICATIONS OF ROUTHHURWITZ CRITERION TO
LINEAR FEED BACK SYSTEMS:
O
The Routh stability criterion is frequently used for the
determination of the condition of stability of linear feedback
systems as shown in the in the following example:
Ex 5):
Consider the closed loop transfer function of the system as
C(s)/R(s)=K/{s*(s
2
+ s + 1)*(s + 4)+K}
Then , the characteristic equation becomes,
s*(s
2
+ s + 1)*(s + 4)+K=0
s
4
+ 5s
3
+ 5s
2
+ 4s + K=0
s
4
s
3
s
0
s
1
s
2
1 5
4 5
K
21/5
84/5 – 5K
21/5
K
:
:
:
:
:
K
Applying RouthHurwitz stability criterion for the above
system:
Since for a stable system, the signs of elements of
the first column of the Routh array should be all positive ,
the condition of system stability requires that,
K > 0 and
(84/55K) > 0
Therefore for stability , K should lie in the range
84/25 > K > 0
When K=84/25 , there will be a zero at the first entry in
the fourth row of the Routh array . This corresponds to the
presence of a pair of symmetrical roots . Therefore
K=84/25, will cause sustained self oscillations in the
closed
loop system .
LIMITATIONS OF ROUTHHURWITZ
CRITERION:
Routh Hurwitz criterion has the following disadvantages :
O
It gives only the number of roots in the right half of the s
plane .
O
It gives no information as regards the values of the roots
and does not distinguish between real and complex roots .
O
It can’t be used for characteristic equations containing
exponential , trigonometric coefficients .
O
It is very complicated . It is time consuming for higher order
systems .
O
It becomes tough to predict marginal stability for bigger
systems .
STATE VARIABLE METHODS
PRESENTED BY
K.RAJASEKHARA REDDY(7837)
BIOTECHNOLOGY
DEFINITIONS:
STATE VARIBLES IN A SYSTEM
system
State
varibles(x1,x2,x3)
‘m’ inputs ‘p’ outputs
STANDARD STATE FORM EQUATION
STANDARD STATE FORM EQUATION
EXAMPLE WITH GENERAL CIRCUIT…
CIRCUIT DIAGRAM
EQUATIONS:
i1+i2+i3=0
i1+i2+c dv/dt =0
L1 di1/dt +i1R1+(eV)=0
L2 di2/dt +i2R2V=0
dV/dt=x1’=(1/c)(i1+i2)
di1/dt=x2’=(1/L1)(i1R1+eV)
di2/dt =x3’=(1/L2)(i2R2V)
STATE EQUATION FROM PREVIOUS
EQUATIONS
x1’ 0 1/c 1/c x1 o
x2’ = 1/L1 –R1/L1 0 x2 + 1/L1 u
x3’ 1/L2 0 R2/L2 x3 0
Consider the voltage and current in R2 as output
variables y1,y2
y=Cx+Du
INPUT STATE OUTPUT EQUATION
y1 0 0 R2 x1
= x2
y2 0 0 1 x3
TRANSFER FUNCTION FROM STATE
VARIABLE MODELS
x’=Ax+Bu
y=Cx+Du
Taking laplace transform on both sides…
S*x(s)x(0)=A*x(s)+B*u(s)
(sIA)*x(s)=B*u(s)
x(s)=(1/(sIA))*B*u(s)
y(s)=c/(sIA)*B*u(s)+D*u(s)
Transfer function=y(s)/x(s)=(c/(sIA))*B+D
WHY WE GO FOR STATE VARIABLE
TECHNIQUE
Due to many demerits of transfer function we go
for state variable method.
DEMERITS OF TRANSFER FUNCTION
Applicable only for linear systems
Analysis of multiple i/p system become tedious
Applicable only for time invarient system
Initial conditions need to be zero
Does not provide any information about the
internal structure of the system
Issues with stability – there are many cases where
the analysis using transfer function gives overall
stable system but the system elements show the
tendency to exceed the specified ratings.
More of hit and trial error method.
AN INTRODUCTION TO P AND
PI CONTROLLERS.
By:SHASHANK KUMAR
¾ B.TECH, MECHANICAL ENGINEERING
ROLL NO. – 07352
WHAT IS A CONTROLLER ?
O
A controller compares the actual value of a system with the
reference input i.e. the desired value , measures the deviation
and then produces a control signal in accordance with the
error signal so as to reduce the deviation to zero or a
minimum value.
O
This action is called the control action.
DIFFERENT TYPES OF CONTROLLER
The different types of controller used today in industrial
applications can be categorized as follows :
1. Twoposition or onoff controllers.
2. Proportional controllers.
3. Integral controllers.
4. Proportionalplusintegral controllers.
5. Proportionalplusderivativecontrollers.
6. Proportionalplusintegralplusderivative controllers.
PROPORTIONAL CONTROLLER
O
A proportional controller is a type of linear feedback control system which follows
the following relation between input and output:
U(t)= ;
here e(t) is the error signal which is the difference between the reference value and
the process variable.
in laplace form, this is
O
The proportional mode adjusts the output signal in direct proportion to the
controller input. The adjustable parameter to be specified is the proportional
controller gain kp.
O
The proportional controller is basically an amplifier with an adjustable gain.
O
An example is a driver driving a car , if he is over speeding he reduces the
accelerator, if he is at a lower speed he accelerates the car.,
A SIMPLE PROPORTIONAL
CONTROLLER
e(s)
c(s)
ADVANTAGES OF THE PROPORTIONAL
CONTROLLER
1. It does not require a very precise analytical
modeling of the system being controlled.
2. It is very simple to implement
DRAWBACKS OF THE PROPORTIONAL
CONTROLLER
1. A proportional controller reduces error but does not
eliminate it (unless the process has naturally integrating
properties).
2. It causes nonzero steady state error.
3. It does not provide any protection against oscillations.
PROPOTIONAL AND INTEGRAL
CONTROLLER
CONTROLLER PART
THE SPECIFICATION OF VARIOUS
CONSTANTS ARE AS FOLLOWS
O
Kp =PROPOTIONAL GAIN
O
Ki =INTEGRAL GAIN
O
E(t) =ERROR FUNCTION
O
Ti= INTEGRAL TIME
THE EQUATION RELATED WITH IT ARE AS FOLLOWS :
M(t)=Kp e(t) + Kp/Ti e(t) dt
TRANSFER FUNCTION
M(s)/E(s)=Kp(1+1/Ti s)
WHERE Kp : PROPOTIONAL GAIN OF THE CONTROLLER
Ti : INTEGRAL TIME
0
t
THE OUTPUT GRAPH OF PI
CONTROLLER
•
ADVANTAGES OF PI CONTROLLER
The integral term of the PI controller causes the steady
state error to be zero for a step input
LIMITATION OF PI CONTROLLER
The problem with using a PI controller is
that it introduces a phaselag. This means
that the phase margin (a measure of
stability) decreases. So careful design
considerations with respect to the gain must
be considered.
APPLICATION OF PI CONTROLLERS
O
THEY ARE USED IN FINDING AIR FLOWS
MOTOR POSITIONING IN ROBOTS ETC
PROPORTIONAL PLUS
DERIVATIVE CONTROLLER
By
Akhilesh Kumar
ProportionalDerivative control is useful for fast response controllers that do
not need a steadystate error of zero. Proportional controllers are fast.
Derivative controllers are fast. The two together is very fast.
1. Proportional action
• Provides an instantaneous response to the control error.
• This is useful for improving the response of a stable system but cannot
control an unstable system by itself.
• The gain is the same for all frequencies leaving the system with a nonzero
steadystate error.
INTRODUCTION:
2. Derivative action:
• Derivative action acts on the derivative or rate of change of the control error.
• This provides a fast response, as opposed to the integral action, but cannot
accommodate constant errors (i.e. the derivative of a constant, nonzero error is 0).
• Derivatives have a phase of +90 degrees leading to a predictive response.However,
derivative control will produce large control signals in response to high frequency
control errors.
Combination of proportional and derivative:
ProportionalDerivative or PD control combines proportional control and
derivative control in parallel.
where
is the proportional control gain and
is the derivative gain or derivative time.
DESIGN:
Kp
Kd s
R(s)
C(s) E(s)
+
_
+
+
Wn
s(s+2*Wn*e)
The block diagram of PD controller is shown below:
Output, C(s)= (Kp + sKd)E(s)
OPAMP REALIZATION OF PD CONTROLLER:
R1
R2
R
R
E’
E
Transfer function of the circuit : E’/E= R2/R1 + R3*C*s
R
R3
C
AN PRACTICAL EXAMPLE:
Consider the figure shown below:
controll
er
J
Ia= constant
Tw
Tm,Bm
B
Br
Rf
Lf
V
The figure shows an electromechanical servo where
Tm = Motor Torque
J = Inertia
Tw = Torque disturbance
V = Viscous friction coefficient
The damper torque on J, due to V, behaves exactly like a derivative control mode in that it
always opposes velocity dB/dt with a strength proportional to dB/dt, making motion less
oscillatory.
Solving the system and considering second order characteristic equation, we get
Wn = √(K/J) and e = (Td/2) * √(K/J)
If Td > 0, motor torque is felt by the load similar to mechanical viscous
damping torque. The derivative control is hence interpreted as pseudofriction
effect.
The block diagram of the given system is as shown below:
Kp
Kc(1+Td*
s)
1/R
f
Kt
1/Js*
s
Kp
Tw
B Br
+
_
+
+
If we change the system by using a proportional controller in the forward path and
adding derivative mode at the controlled variable B.
And comparing derivative control applied to system error with that to the controlled
variable. We find the similar effect on characteristic equation. However , derivative
control has more violent response to sudden changes due to derivative term Br’s
presence.
If we study the effect of derivative control on steady state errors (say for a unit step
disturbance with Br= zero, we conclude that :
• The performance is unaffected by the derivative mode.
• However, there is an indirect method for gaining improvements in steady state accuracy
i.e. by proper choice of derivative mode allows the use of higher values of proportional
gain, thereby improving the steady state accuracy.
EFFECTS OF PD CONTROL:
1. Improving damping and reducing maximum
overshoot.
2. Reducing rise time and settling time.
3. Increasing Bandwidth.
4. Possibly accentuating noise at higher
frequencies.
5. Possibly requiring a relatively large capacity in
implementation.
APPLICATIONS:
1. Fast steering mirror.
2. Helicopter Hover.
3. Inverted pendulum.
4. Mixing tank.
5. Vibration Isolation.
6. Simple offline least squares system identification.
ROOT LOCUS
(THEORY)
Consider the following standard control system,
C(s)
G(s)
H(s)
R(s)
error
1+ KG(s)H(s)
Where K is varied
• Transience and stability of a system depends upon its CLOSED LOOP
POLES.
•To build a better system, movement of poles can be adjusted by modifying
the system parameters.
•The ROOT LOCUS method thus is of help.
What is a ROOT LOCUS?
¬
Plot of the LOCI of the closed loop poles as a function of the open loop
gain (K), where K is varied from –infinity to infinity
When,
¬K is varied from –infinity to zero, it is a direct root locus.
¬K is varied from zero to infinity, it is a inverse root locus.
There are 8 steps to constructing a ROOT LOCUS and must be followed
in order to reach a conclusion
RULE 1:
• Root locus is always symmetrical to the xaxis or the real axis.
• Roots are either real or complex conjugate pairs or both.
• Hence the roots or poles and zeroes of the equation are found first.
Say,
G(s)=_________(s+1)___
(s^2 + 4s +5)(s+3)
So the poles of the equation are
s+3=0, s=3
s^2 +4s+5=0 s=2+I and 2I
And the poles are
s+1=0 s=1
CONSTRUCTION OF A ROOT LOCUS
RULE 2: Number of LOCI
•If the number of poles be n and the number of zeroes be m.
•If n>m , then
Total number of loci= n
•Each locus starts from a pole and ends in a zero.
•So, number of loci that start from a pole and end in infinity is nm.
E.g. in the previous case
Total number of loci=3
And number of loci starting from a pole and end in infinity is 32=1
Rule 3: Real axis loci
• Some of the loci will lie on the real axis.
• A point on the real axis will lie on the root locus if and only if,
Number of poles and zeroes to the right is odd
RULE 4: Angle of asymptotes
•Mostly, number of poles > number of zeroes. i.e. n>m
Hence, nm branches move to infinity, along the asymptote.
Asymptote: It is defined as a line on the root locus which
converges at infinity.
The angle of asymptote is given by,
Theta= (2*q + 1)*180 / nm for K>0
Theta= (2*q)*180 / nm for K<0
Where q=0,1,2….nm1
e.g. Using the same example as before we see that,
Q can take 0,1 as the values. Nm=2
Theta=90, 270 degrees.
RULE 5: Centre of asymptotes
• Since knowledge of only asymptote angle is insufficient, the location of
asymptotes in the splane are also important
• The point where the asymptote touches the real axis is known as the centroid.
•Centroid= [sum(real part of poles) sum(real parts of zeroes)]/(nm)
e.g. In the above example
G(s)=_________(s+1)___
(s^2 + 4s +5)(s+3)
sum(real part of poles) = 2+(2)+(3)= 7
sum(real parts of zeroes) = 1
nm =2
•Centroid= (7+1)/2= 3
RULE 6: break away and breakin point
• Breakin point is defined as the point where the root locus enters real axis.
• Break away point is defined as the point where the root locus comes out of the real
axis.
•Break away & break in points are points on the real axis at which multiple roots of
the characteristic equation occur.
•Observation: If there are 2 adjacently placed poles on the real axis and the real axis
is a part of the root locus, hence minimum 1 break away point exists between the 2
poles.
1+ G(s)H(s)=0; we get a relation in terms of K.
dk/ds=0, gives the break away point.
e.g. G(s)H(s)=K / s^2 + 4s+5
Hence,
s^2+4s+5+K=0, and K=(s^2+4s+5)
dk/ds=0, ¬ 2s+4=0
hence, s=2.
RULE 7: Intersection of root locus with jw axis.
• To calculate the intersection with jw axis, this is followed
1.Construct 1+G(s)H(s)=0
2.Develop the routh array in terms of K
3.Find K value for which routh array contains a term of zeroes.
4.Frame auxiliary equation with the help of coefficients from the row above the row of
zeroes.
5.Substitute the K value found in step 2, equate the equation to zero, find the value of
s.
The roots of the equation give the intersection points.
e.g. say G(s)H(s)= K/s*s+1*s+3
Characteristic equation will be s^3+ 4s^2 +3s+K=0,
Routh array s^3: 1 3
s : 4 K
s^0: (12K)/4…..hence K12=0 ¬ K=12
Substituting K=12 in the auxiliary equation, 4s^2+K=0,
hence we get K=(+or) root(3)i , which are the intersection points.
RULE 8: Angle of departure and angle of arrival.
•The root locus always leaves a complex pole, at an angle known as the departure
angle given by,
Theta=180+ arg( G(s)H(s) )
•The root locus always arrives at a complex zero, at an angle known as the arrival
angle.
Theta=180 arg( G(s)H(s) ),
Where G(s)H(s) is the angle excluding the zero where the angle has the be
calculated
Steps of a ROOT LOCUS method:
1.Determine the branch number ending at infinity of the loci using rule 1.
2.Plot the poles and zeroes.
3.Find the real axis loci using dark lines.
4.Find the asymptotes and their angles using rule 3.
5.Using rule 4 determine the asymptotes centre and draw step 4 and step 5.
6.Calculate the break in or breakaway points.
7.Calculate angle of departure and angle of arrival using the rule.
8.Determine the jw crossover if root locus has complex poles and zeroes.
By
K S Venkatakrishnan
7328
Mechanical Engineering
FREQUENCY
FREQUENCY
DOMAIN ANALYSIS
DOMAIN ANALYSIS
BY
T V D .GOWTHAM
ROLL NO 7356
¾ B.TECH
MECHANICAL ENGINEERING
FREQUENCY RESPONSE
×
The term frequency response means the steady
state response of a system to a sinusoidal input
×
In frequency response analysis,we vary the
frequency of input signal over a certain range
and study the resulting response
O
Frequency response requirements differ
depending on the application.
O
Frequency response curves are often used to
indicate the accuracy of electronic components
or systems
WHY FREQUENCY RESPONSE ANALYSIS
COMES INTO PICTURE
×
The frequency response of a control system
presents a qualitative picture of the transient
response
×
the frequency response allows us to understand
a circuit's response to more complex inputs
×
Time domain analysis is limited to only smaller
systems,the stability of larger systems is difficult
to analyse
×
Generation of sinusoidal signals is very easy
×
Frequency response measurements can be used
directly to quantify system performance and
design control systems.
FREQUENCY RESPONSE FOR A GIVEN
SINUSOIDAL INPUT
×
The sinusoidal transfer function G(jw) ,the ratio of
Y(jw) to X(jw),is a complex quantity and can be
represented by magnitude and phase angle with
frequency as a parameter
×
A negative phase angle is called phase lag and
positive phase angle is called phase lead
×
The sinusoidal transfer function of linear system
can be obtained by substituting jw in S of a
transfer function
REPRESENTATION OF FREQUENCY RESPONSE
CHARACTERISTICS IN GRAPHICAL FORMS
+
There are three commonly used representations of
sinusoidal transfer functions:
×
Bode plots
×
Polar plots
×
Nyquist plots
1. INTRODUCTION
2. ILLUSTRATION
3. GUIDELINES FOR A POLAR PLOT
4. SIMPLE PLOTS
5. STABILITY OF POLAR PLOT
6. ADVANTAGES OF FREQUENCY
RESPONSE OVER TIME RESPONSE
ANALYSIS
POLAR PLOT
A POLAR PLOT IS DEFINED AS THE LOCUS OF GH(JW) PHASOR
IN POLAR REPRESENTATION WHEN ‘W’ CHANGES FROM INFINITY
TO ZERO AND THE PLOT IS IN GH(JW) PLANE
O
Here the plot is between IG(jw)I & IG(jw) as a function of
w
O
This is used in determining the stability of closed loop
system from it’s open loop frequency response .For closed
loop system stability, open loop transfer function is G(s)
H(s) instead of G(s) unless H(s) is 1.Unless stated above
the open loop tranfer function is G(s).
ILLUSTRATION OF POLAR PLOTS
G(s)H(s) = K (s+a)/s(s+b) (s+c)
Replacing s by jw ,we have
G(jw)H(jw)= K(a+jw)/jw(b+jw)(c+jw)
= lGH(jw)l F(w) is called polar representation
lGH(jw)l is called magnitude and F(w) is called phase angle
lGH(jw)l = K(a^2+w^2)^.5 /jw(b^2+w^2)^.5 (c^2+w^2)^.5;
F(w) = arg(a+jw)arg(b+jw)arg(c+jw)90
Let us suppose GH(s) =K/(s+4)(s+6)(s+9)
putting s =jw
GH(jw) = K/(jw+4)(jw+6)(jw+9)
= K/{(16+w^2)^.5 arg(4+jw) (36+w^2)^.5
arg(6+jw) (81+w^2)^.5 arg(9+jw)}
Next step is to variation of w from infinity to 0
GUDELINES TO DRAW THE POLAR PLOTS
1.IDENTIFY G(s)H(s)
2.APPLY W TENDS TO INFINITY BY CONSIDERING
ONLY HIGHER ORDER TERMS
3.APPLY W TENDS TO ZERO BY CONSIDERING ONLY
LOWER ORDER TERMS
4.CONSIDER W=1 AS A TRAIL POINT AND GIVE THE
LOCATION OF ITS MAGNITUDE AND PHASE ANGLE
5.NOW GIVE THE APPROXIMATE SHAPE OF POLAR
PLOT IN GH(w) PLANE
6.APPLY POLAR PLOT ANALYSIS TO IDENTIFY THE
OBSERVED STABILITY AND DEGREE OF STABILITY
OF CONTROL SYSTEM
FOR G(s) = K/[(s+4)(s+6)(s+9)] POLAR PLOT IS
Similarly for the following we get the corresponding graphs in polar plane
and XY PLANE
. G(s)H(s)= (1+sT)^1
SIMPLE PLOTS OF THE FOLLOWING FUNCTIONS
Polar plot of G(s)=1/(S+2)
First substitute s = jω in G(s)
2
tan ) (
4
1
) (
1
2
ω
ω
ω
ω
−
· ∠
+
· j G j G
polar plot for the system with
) 2 )( 1 (
10
) (
+ +
·
s s s
s G
) 2 )( 1 (
10
) (
+ +
·
ω ω ω
ω
j j j
j G
) 2 ( ) 1 (
10
) (
2 2
+ +
·
ω ω ω
ω j G
ω
ω
ω
1 1
tan
2
tan 90 ) (
− −
− − − · ∠
j G
The plot is shown in Fig
STABILITY ANALYSIS OF POLAR PLOTS
HERE WE ARE CONSIDERING CHARACTERISTIC EQUATION
1+GH(jw) = 0 ‘ WHERE (1,J0) IS CONSIDERED AS ORIGIN AND AS
A ‘CRITICAL POINT’ W.R.T CLOSED LOOP STABILITY
º
IF POLAR PLOT PASSES THROUGH (1,J0) POINT SYSTEM IS
CRITICALLY STABLE
º
IF (1,JO)IS NOT ENCLOSED BY PLOT SYSTEM IS ABSOLUTELY
STABLE
º
IF (1,J0) IS ENCLOSED BY THE PLOT THEN THE SYSTEMIS
UNSTABLE
CONCEPT OF ENCLOSURE : A POINT OR REGION IS SAID TO BE
ENCLOSED BY POLAR PLOT IF THE POINT OR REGION LIES ON
THE LHS OF DIRECTION OF JOURNEY OF PLOT STARTING FROM
W =INFINITY TO W=0
FOR FINDING OUT STABILITY
Q
CONSIDER THE EXPRESSION AND RATIONALISE IT
Q
EQUATE THE IMAGINARY VALUE T0 ZERO TO FIND OUT PHASE
CROSSOVER FRQUENCY FIND OUT GH(jw) AND EQUATE IT TO a+JO
Q
SUPPOSING CRITICAL K=SOME VALUE WE CALCULATE ‘a’
Q
IT IS SAID
lal = 1 SYSTEM IS CRITICALLY STABLE
lal < 1 SYSTEM IS STABLE
lal >1 SYSTEMIS UNSTABLE
WHERE a=(CRITICAL K)/(ACTUAL K)
GAIN MARGIN ALSO DETERMINES THE STABILITY : GM=20 loglal
+
GM= 6dB IS THE BEST RANGE
+
GM=0 Db CRITICALLY STABLE
+
GM= 6dB UNSTABLE
HENCE POLAR PLOT AND CONCEPT OF ENCIRCLEMENT ARE ABLE TO
PREDICT THE STABILITY OF A SYSTEM.
FURTHER THIS PORTION OF POLAR PLOT IS USED IN NYQUIST CRITERION
PRESENTATION ON BODE PLOTS
By
Jaivee T.Joseph
O7324
Mechanical dept.
¾ B. Tech
BODE PLOTS
A Bode plot is a graph of the logarithm of the transfer
function of a linear, timeinvariant system versus
frequency, plotted with a logfrequency axis, to show the
system's frequency response. It is usually a combination
of a Bode magnitude plot (usually expressed as dB of
gain) and a Bode phase plot (the phase is the imaginary
part of the complex logarithm of the complex transfer
function).
EXPRESSING IN DB
Given the tranfer function:
) 1 / )( (
) 1 / (
) (
+
+
·
p jw jw
z jw K
jw G
B
20logG(jw) = 20logK + 20log(jw/z+1)  20logjw 
20logjw/p + 1
The gain term, 20logK
B,
is just so many
dB and this is a straight line on Bode paper,
independent of omega (radian frequency).
The term,  20logjw =  20logw, when plotted
on semilog paper is a straight line sloping at

20dB/decade. It has a magnitude of 0 at w = 1.
0
20
20
ω =1
20db/dec
wlg
Poles, zeros and bode plots
The term,  20log(jw/p + 1), is drawn with the
following approximation: If w < p we use the
approximation that –20log(jw/p + 1 ) = 0 dB,
a flat line on the Bode. If w > p we use the
approximation of –20log(w/p), which slopes at
20dB/dec starting at w = p. Illustrated below.
It is easy to show that the plot has an error of
3dB at w = p and – 1 dB at w = p/2 and w = 2p.
One can easily make these corrections if it is
appropriate.
0
20
20
40
ω = p
20db/dec
wlg
0
20
20
40
ω = z
+20db/dec
When we have a term of 20log(jw/z + 1) we
approximate it be a straight line of slop 0 dB/dec
when w < z. We approximate it as 20log(w/z)
when w > z, which is a straight line on Bode paper
with a slope of + 20dB/dec. Illustrated below.
wlg
STRAIGHTLINE AMPLITUDE PLOT
At every value of s where ω = x
n
(a zero), increase
the slope of the line by per decade.
At every value of s where ω = y
n
(a pole), decrease
the slope of the line by per decade.
The initial value of the graph depends on the
boundaries. The initial point is found by putting the
initial angular frequency ω into the function and
finding H(jω).
The initial slope of the function at the initial value
depends on the number and order of zeros and poles
that are at values below the initial value, and are
found using the first two rules.
STRAIGHTLINE PHASE PLOT
To draw the phase plot, for each pole and zero:
if A is positive, start line (with zero slope) at 0 degrees
if A is negative, start line (with zero slope) at 180 degrees
at every ω = x
n
(for stable zeros – Re(z) < 0), increase the slope by
degrees per decade, beginning one decade before ω = x
n
(E.g.: )
at every ω = y
n
(for stable poles – Re(p) < 0), decrease the slope by
degrees per decade, beginning one decade before ω = y
n
(E.g.: )
"unstable" (right half plane) poles and zeros (Re(s) > 0) have
opposite behavior
flatten the slope again when the phase has changed by degrees
(for a zero) or degrees (for a pole),
After plotting one line for each pole or zero, add the lines together to
obtain the final phase plot; that is, the final phase plot is the
superposition of each earlier phase plot.
ADVANTAGES
O
In the absence of a computer, a bode plot
can be sketched by approximating the
magnitude and phase with right line
segments.
O
Gain crossover, phase crossover, gain
margin and phase margin are more easily
determined on the bode plot than the other
plots
O
For design purposes, the affect of adding
controllers and their parameters are more
easily visualized here.
DISADVANTAGES
O
Absolute and relative stability of only minimum
phase systems can be determined from bode
plots
O
They are useful only for stability studies of
systems with minimum phase loop transfer
functions
ADVANTAGES OF FREQUENCY RESPONSE OVER TIME
RESPONSE ANALYSIS
×
Here a system may be designed so that effects of undesirable noises are
negligible and such analysis and design canbe extended to certain
nonlinear systems
×
In design problems, there are no unified methods of arriving at a
designated system that meets timedomain performance specifications. On
the other hand, in frequency domain, a wealth of graphical and other
techniques are available that are useful for system analysis and design,
irrespective of the order of the system
×
We can use the data obtained from measurements on the physical system
without deriving its mathematical model
×
Frequency response tests are simple and canmake use readily available
sinusoidal signal generators
DIGITAL CONTROL
SYSTEM
Ashique Poyilil
7308
Mechanical engineering
238
O
Digital Control System  (DCS) A digital computer used
for realtime control of a dynamic system, usually in an
industrial environment, possibly as part of a
Supervisory Control and Data Acquisition (SCADA)
system.
O
A DCS samples feedback from the system under
control and modifies the control signals in an attempt
to achieve some desired behaviour.
239
O
Analysis of such digitalanalogue feedback
systems can involve mathematical methods such
as difference equations, Laplace transforms, z
transfer functions, state space models and state
transition matrices.
240
DIGITAL/DISCRETE CONTROL
More useful for computer systems
Time is discrete
denoted k instead of t
Main tool is ztransform
f(k) → F(z) , where z is complex
Analogous to Laplace transform for sdomain
Rootlocus analysis has similar flavor
Insights are slightly differen
241
∑
∞
·
−
· ·
0
) ( ) ( )] ( [
k
k
z k f z F k f Z
OUTLINE
Introduction to digital control
Digitization
Effect of sampling
Sampling
Spectrum of a sampled signals
Sampling theorem
Discrete Systems
Ztransform
Transfer function
Pulse response
Stability
242
DIGITIZATION
O
Analog Control System
243
ctrl. filter
D(s)
plant
G(s)
sensor
1
r(t) u(t) y(t) e(t)
+

continuous controller
For example, PID control
DIGITIZATION
O
Digital Control System
O
T is the sample time (s)
O
Sampled signal : x(kT) = x(k)
244
control:
difference
equations
D/A and
hold
sensor
1
r(t) u(kT) u(t) e(kT)
+

r(kT)
plant
G(s)
y(t)
clock
A/D
T
T
y(kT)
digital controller
voltage → bit
bit → voltage
DIGITIZATION
O
Continuous control vs. digital control
¤
Basically, we want to simulate the cont. filter D(s)
¤
D(s) contains differential equations (time domain) – must be
translated into difference equations.
O
Derivatives are approximated (Euler’s method)
245
T
k x k x
k x
) ( ) 1 (
) (
− +
≈
`
DIGITIZATION
246
Example (3.1)
Using Euler’s method, find the difference equations.
b s
a s
K
s E
s U
s D
+
+
· ·
0
) (
) (
) (
Differential equation
) ( ) ( ) ( ) ( ) (
0 0
1
ae e K bu u s E a s K s U b s
L
+ · + ÷→ ÷ + · +
−
` `
Using Euler’s method
) 1 ( ) ( ) 1 ( ) ( ) 1 ( ) 1 (
) ( ) 1 ( ) ( ) 1 (
0 0
0
+ + − + − · +
⇒
,
`
.

+
− +
· +
− +
k e K k e aT K k u bT k u
ae
T
k e k e
K bu
T
k u k u
DIGITIZATION
247
s s
s s
s G
s
s
s D
+
·
+
·
+
+
·
2
1
) 1 (
1
) ( ,
10
2
70 ) (
Compare – investigate using Matlab
1) Closed loop step response with continuous controller.
2) Closed loop step response with discrete controller.
Sample rate = 20 Hz
3) Closed loop step response with discrete controller.
Sample rate = 40 Hz
Significance of sampling time T
Example controller D(s) and plant G(s)
DIGITIZATION
248
Matlab  continuous controller
numD = 70*[1 2]; denD = [1 10];
numG = 1; denG = [1 1 0];
sysOL = tf(numD,denD) * tf(numG,denG);
sysCL = feedback(sysOL,1);
step(sysCL);
Matlab  discrete controller
numD = 70*[1 2]; denD = [1 10];
sysDd = c2d(tf(numD,denD),T);
numG = 1; denG = [1 1 0];
sysOL = sysDd * tf(numG,denG);
sysCL = feedback(sysOL,1);
step(sysCL);
s s
s G
s
s
s D
+
·
+
+
·
2
1
) (
10
2
70 ) (
Controller D(s)
and plant G(s)
DIGITIZATION
249
Notice, high sample frequency (small sample time T )
gives a good approximation to the continuous controller
EFFECT OF SAMPLING
250
D/A in output from controller
The single most important impact of implementing a control
digitally is the delay associated with the hold.
EFFECT OF SAMPLING
Analysis
Approximately 1/2 sample time delay
Can be approx. by Padè
(and cont. analysis as usual)
251
ctrl. filter
D(s)
Padé
P(s)
sensor
1
r(t) u(t) y(t) e(t)
+

plant
G(s)
T s
T
s P
/ 2
/ 2
) (
+
·
2
T
T
d
·
EFFECT OF SAMPLING
252
Example of phase
lag by sampling
Example from
before with sample
rate = 10 Hz
Notice PM reduction
SPECTRUM OF A SAMPLED SIGNAL
Spectrum
Consider a cont. signal r(t)
with sampled signal r*(t)
Laplace transform R*(s) can be
calculated
253
∑
∑
∞
−∞ ·
∗
∞
−∞ ·
∗
− ·
− ·
k
s
k
jn s R
T
s R
kT t t r t r
) (
1
) (
) ( ) ( ) (
ω
δ
r(t) r*(t)
T
SPECTRUM OF A SAMPLED SIGNAL
254
SPECTRUM OF A SAMPLED SIGNAL
O
High frequency signal and low frequency signal – same
digital representation.
255
SPECTRUM OF A SAMPLED SIGNAL
O
Removing (unnecessary) high frequencies – anti
aliasing filter
256
control:
difference
equations
D/A and
hold
sensor
1
r(t) u(kT) u(t) e(kT)
+

r(kT)
plant
G(s)
y(t)
clock
A/D
T
T
y(kT)
digital controller
anti
aliasing
filter
SPECTRUM OF A SAMPLED SIGNAL
257
SAMPLING THEOREM
O
Nyquist sampling theorem
¤
One can recover a signal from its samples if the
sampling frequency f
s
=1/T (ω
s
=2π /T) is at least
twice the highest frequency in the signal, i.e.
¤
ω
s
> 2 ω
b
(closed loop bandwidth)
O
In practice, we need
¤
20 ω
b
< ω
s
< 40 ω
b
258
DISCRETE SYSTEMS
A discrete system is a system with a countable number of
states.
Discrete systems may be contrasted with continuous
systems, which may also be called analog systems.
A discrete system is often modeled with a directed
graph (mathematics) and is analyzed for correctness and
complexity according to computational theory. Because
discrete systems have a countable number of states, they
may be described in precise mathematical models.
259
O
A computer is a finite state machine that may be
viewed as a discrete system.
O
Because computers are often used to model not
only other discrete systems but continuous
systems as well, methods have been developed
to represent realworld continuous systems as
discrete systems. One such method involves
sampling a continuous signal at discrete time
intervals.
260
ZTRANSFORMS
ARUN.K
7307
ZTRANSFORMS
×
The Ztransform, like many integral transforms,
can be defined as either a onesided or twosided
transform
×
The ztransform is the most general concept for
the transformation of discretetime series.
×
The Laplace transform is the more general concept
for the transformation of continuous time
processes.
×
For example, the Laplace transform allows you to
transform a differential equation, and its
corresponding initial and boundary value problems,
into a space in which the equation can be solved by
ordinary algebra.
×
The switching of spaces to transform calculus
problems into algebraic operations on transforms is
called operational calculus. The Laplace and z
transforms are the most important methods for this
purpose.
CS851: Feedback Control for Computer Systems
Control System
Controller Target
System
Transducer
Reference
Input
Control
error
Control
Input
Measured
Output
Transduced
Output
264
The Transforms
×
The Laplace transform of a function f(t):
∫
∞
−
·
0
) ( ) ( d t e t f s F
s t
×
The onesided ztransform of a function x(n):
∑
∞
·
−
·
0
) ( ) (
n
n
z n x z X
×
The twosided ztransform of a function x(n):
∑
∞
−∞ ·
−
·
n
n
z n x z X ) ( ) (
265
Relationship to Fourier Transform
Note that expressing the complex variable z in polar form reveals the
relationship to the Fourier transform:
∑
∑
∑
∞
−∞ ·
−
∞
−∞ ·
− −
−
∞
−∞ ·
· ·
· ·
·
n
n i i
n
n i n i
n
n
i i
e n x X e X
r if and e r n x re X
or re n x re X
ω ω
ω ω
ω ω
ω ) ( ) ( ) (
, 1 , ) ( ) (
, ) )( ( ) (
which is the Fourier transform of x(n).
×
The Ztransform is a generalization of the
discretetime Fourier transform (DTFT).
×
The DTFT can be found by evaluating the
Ztransform at or, in other words, evaluated
on the unit circle. In order to determine the
frequency response of the system the Z
transform must be evaluated on the unit circle,
meaning that the system's region of
convergence must contain the unit circle
266
Region of Convergence
×
The ztransform of x(n) can be viewed as the Fourier transform of x(n)
multiplied by an exponential sequence r
n
, and the ztransform may
converge even when the Fourier transform does not.
×
By redefining convergence, it is possible that the Fourier transform may
converge when the ztransform does not.
×
For the Fourier transform to converge, the sequence must have finite
energy, or:
∞ <
∑
∞
−∞ ·
−
n
n
r n x ) (
Continued,
267
∑
∞
−∞ ·
−
·
n
n
z n x z X ) ( ) (
The power series for the ztransform is called a Laurent series:
The Laurent series, and therefore the ztransform, represents an analytic
function at every point inside the region of convergence, and therefore the z
transform and all its derivatives must be continuous functions of z inside the
region of convergence.
In general, the Laurent series will converge in an annular region of the z
plane.
Example (causal ROC),
ROC shown in blue, the unit circle as a dotted grey circle and the circle is
shown as a dashed black circle
Let (where u is the Heaviside step function). Expanding on the interval it
becomes
X[n]={….,0,0,0,1,0.5,0.5^2,0.5^3,…..}
269
Poles and Zeros
When X(z) is a rational function, i.e., a ration of polynomials in
z, then:
1. The roots of the numerator polynomial are referred to as
the zeros of X(z), and
2. The roots of the denominator polynomial are referred to as
the poles of X(z).
Note that no poles of X(z) can occur within the region of convergence since the
ztransform does not converge at a pole.
Furthermore, the region of convergence is bounded by poles.
270
Properties
•
ztransforms are linear:
•
The transform of a shifted sequence:
•
Multiplication:
But multiplication will affect the region
of convergence and all the polezero
locations will be scaled by a factor of
a.
[ ] ) ( ) ( ) ( ) ( z bY z aX n by n ax + · + Z
[ ] ) ( ) (
0
0
z X z n n x
n
· + Z
[ ] ) ( ) (
1
z a Z n x a
n −
· Z
ZTRANSFORMS OF COMMON
FUNCTIONS
271
Name f(t) F(z)
Impulse
Step
Ramp
Exponential
Sine
1
1 − z
z
2
) 1 ( − z
z
a
e z
z
−
1 ) (Cos 2
Sin
2
+ − z a z
a z
1 ) ( · t f
t t f · ) (
at
e t f · ) (
) sin( ) ( t t f ω ·
F(s)
1
s
1
2
1
s
a s −
1
2 2
1
s + ω
¹
'
¹
>
·
·
0 0
0 1
) (
t
t
t f
D.C MOTOR SPEED
CONTROL
BANDI SREENIHAR
By
º
INTRODUCTION
º
MODELLING
º
PID CONTROLLER
º
TIME RESPONSE AND ROOT LOCUS
º
FREQUENCY RESPONSE AND BODE
PLOT
º
APPLICATIONS
WHAT IS A D.C MOTOR ??
O
MOTOR: Which converts electrical
energy into mechanical energy ,mainly
into speed.
O
And a D.C MOTOR is one which uses
DIRECT CURRENT to convert electrical
energy into mechanical energy.
INTRODUCTION
The electric motors are most widely used as
energy converters in the
O
ROBOTS
O
MACHINE TOOLS
These require AUTOMATIC control of their
main parameters like position, speed,
acceleration and currents.
MAIN PURPOSE OF D.C MOTOR
SPEED CONTROL
O
The purpose of a motor speed controller is to take a signal representing the
demanded speed, and to drive a motor at that speed.
O
The controller may or may not actually measure the speed of the motor.
O
If it does, it is called a Feedback Speed Controller or Closed Loop Speed
Controller,
O
if not it is called an Open Loop Speed Controller.
There are three major types of energy
converters.
O
ELECTRIC MOTORS
O
PNEUMATIC MOTORS
O
HYDRAULIC MOTORS
Electric motors can be grouped into
four major classes.
O
DC motors
O
Stepper motors
O
Asynchronous motors
O
Synchronous motors.
MODELING
Physical setup and system equations:
A common actuator in control systems is the
DC motor. It directly provides rotary
motion and, coupled with wheels or drums
and cables, can provide transitional motion.
The electric circuit of the armature and the
free body diagram of the rotor are shown in
the following figure:
CIRCUIT DIAGRAM
¤
For this system, we will assume the following values for
the physical parameters.
¤
* moment of inertia of the rotor (J) = 0.01 kg.m^2/s^2
* damping ratio of the mechanical system (b) = 0.1 Nms
* electromotive force constant (K=Ke=Kt) = 0.01 Nm/Amp
* electric resistance (R) = 1 ohm
* electric inductance (L) = 0.5 H
* input (V): Source Voltage
* output (theta): position of shaft
¤
The rotor and shaft are assumed to be rigid
The motor torque, T, is related to the
armature current, i, by a constant
factor Kt. The back emf, e, is related
to the rotational velocity by the
following equations:
¤
In SI units (which we will use), Kt (armature constant) is
equal to Ke (motor constant).
¤
From the figure (from above slide) we can write the
following equations based on Newton's law combined with
Kirchoff's law:
O
Transfer Function
O
Using Laplace Transforms, the above modeling equations
can be expressed in terms of s.
By eliminating I(s) we can get the
following openloop transfer
function, where the rotational speed
is the output and the voltage is the
input.
DESIGN REQUIREMENTS
O
Our uncompensated motor can only rotate at
1 rad/sec with an input voltage of 1 Volt .
Since the most basic requirement of a motor is
that
O
It should rotate at the desired speed
O
The steadystate error of the motor speed
should be less than 1%.
O
The other performance requirement is that the
motor must accelerate to its steadystate speed
as soon as it turns on.
We have chosen our design requirements.
We simulate the reference input (r) by an unit step
input, then the motor speed output should have:
O
Settling time less than 2 seconds
O
Overshoot less than 5%
O
Steadystate error less than 1%
MATLAB REPRESENTATION AND OPENLOOP
RESPONSE
1. Transfer Function
J=0.01;
b=0.1;
K=0.01;
R=1;
L=0.5;
num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
step(num,den,0:0.1:3)
title('Step Response for the Open Loop System')
From the plot we see that when 1 volt
is applied to the system, the motor
can only achieve a maximum speed
of 0.1 rad/sec, ten times smaller than
our desired speed. Also, it takes the
motor 3 seconds to reach its steady
state speed; this does not satisfy our 2
seconds settling time criterion.
O
There are several well known methods to
control DC motors such as:
O
PI
O
PID or
bipositional.
PID DESIGN METHOD FOR DC MOTOR SPEED
CONTROL
When we use a controller the schematic
system looks like this.
PROPORTIONAL CONTROL
Let's first try using a proportional controller with a GAIN of 100.
J=0.01;
b=0.1;
K=0.01;
R=1;
L=0.5;
num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
Kp=100;
numa=Kp*num;
dena=den;
[numac,denac]=cloop(numa,dena);
t=0:0.01:5;
step(numac,denac,t)
title('Step response with Proportion Control')
¤
From the plot above we see that both the
steadystate error and the overshoot are
too large.
¤
So lets add an integral term which
eliminates the steadystate error and a
derivative term which will reduce the
overshoot by introducing a PID
controller with small Ki and Kd.
The TRANSFER FUNCTION
for a PID CONTROLLER
J=0.01;
b=0.1;
K=0.01;
R=1;
L=0.5;
num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
Kp=100;
Ki=1;
Kd=1;
numc=[Kd, Kp, Ki];
denc=[1 0];
numa=conv(num,numc);
dena=conv(den,denc);
[numac,denac]=cloop(numa,dena);
step(numac,denac)
title('PID Control with small Ki and Kd')
PUTTING KI=200
PUTTING KD=10
O
So now we know that if we use a PID
controller with
O
Kp=100,
Ki=200,
Kd=10,
O
all of our design requirements will be
satisfied.
ROOT LOCUS DESIGN METHOD FOR DC MOTOR
SPEED CONTROL
Our system schematic looks like this.
DRAWING THE OPENLOOP ROOT LOCUS
The main idea of root locus design is to find
the closedloop response from the openloop
root locus plot. Then by adding zeros and/or
poles to the original plant, the closedloop
response can be modified. Let's first view the
root locus for the plant.
J=0.01;
b=0.1;
K=0.01;
R=1;
L=0.5;
num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
rlocus(num,den)
sgrid(.8,0)
sigrid(2.3)
title('Root Locus without a controller')
FINDING THE GAIN USING THE RLOCFIND
COMMAND
If you recall, we need the settling time and the overshoot to be
as small as possible. Large damping corresponds to points on
the root locus near the real axis. A fast response corresponds
to points on the root locus far to the left of the imaginary
axis. To find the gain corresponding to a point on the root
locus, we can use the rlocfind command. We can find the
gain and plot the step response using this gain all at once.
rlocfind(num,den)
[numc,denc]=cloop(k*num,den,1);
t=0:0.01:3; step(numc,denc,t)
title('Step [k,poles response with gain')
J=0.01; J=0.01;
b=0.1; b=0.1;
K=0.01; K=0.01;
R=1; R=1;
L=0.5; L=0.5;
num=K; num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)]; den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
As we can see, the system is over damped and
the settling time is about one second, so the
overshoot and settling time requirements are
satisfied. The only problem we can see from
this plot is the steadystate error of about
50%. If we increase the gain to reduce the
steadystate error, the overshoot becomes too
large.
FREQUENCY RESPONSE
ANALYSIS
The frequency response method may be less intuitive than other
methods. However, it has certain advantages, especially in real
life situations such as modelling transfer functions from physical
data. The frequency response is a representation of the system's
response to sinusoidal inputs at varying frequencies. The output
of a linear system to a sinusoidal input is a sinusoid of the same
frequency but with a different magnitude and phase. The
frequency response is defined as the magnitude and phase
differences between the input and output sinusoids. In this
presentation, we will see how we can use the openloop frequency
response of a system to predict its behaviour in closedloop.
To plot the frequency response, we create a vector of frequencies
(varying between zero or "DC" and infinity) and compute the value
of the plant transfer function at those frequencies. If G(s) is the open
loop transfer function of a system and w is the frequency vector, we
then plot G(j*w) vs. w. Since G(j*w) is a complex number, we can
plot its magnitude and phase.
“THE BODE PLOT”.
BODE PLOTS
A Bode plot is the representation of the magnitude and phase of G(j*w)
(where the frequency vector w contains only positive frequencies).
*Axes of the figure: The frequency is on a logarithmic scale, the phase
is given in degrees, and the magnitude is given as the gain in
decibels.
A decibel is defined as 20*log10 ( G(j*w )
GAIN AND PHASE MARGIN
Let's say that we have the following system:
where K is a variable (constant) gain and G(s) is the plant under
consideration.
Gain margin: is defined as the change in open loop gain required to
make the system unstable. Systems with greater gain margins can
withstand greater changes in system parameters before becoming
unstable in closed loop. We should keep in mind that unity gain in
magnitude is equal to a gain of zero in dB.
Phase margin: is defined as the change in open loop phase shift
required to make a closed loop system unstable.
The phase margin is the difference in phase
between the phase curve and 180 deg at the
point corresponding to the frequency that
gives us a gain of 0dB (the gain cross over
frequency, Wgc). Likewise, the gain margin is
the difference between the magnitude curve
and 0dB at the point corresponding to the
frequency that gives us a phase of 180 deg
(the phase cross over frequency, Wpc).
The phase margin also measures the system's tolerance to time delay. If
there is a time delay greater than 180/Wpc in the loop (where Wpc is
the frequency where the phase shift is 180 deg), the system will
become unstable in closed loop. The time delay can be thought of as an
extra block in the forward path of the block diagram that adds phase to
the system but has no effect on the gain. That is, a time delay can be
represented as a block with magnitude of 1 and phase w*time_delay
(in radians/second).
We can find the gain and phase margins for a
system directly, by using Matlab by just
entering the margin command. This command
returns the gain and phase margins, the gain
and phase cross over frequencies, and a
graphical representation of these on the Bode
plot.
FREQUENCY DESIGN METHOD FOR DC
MOTOR SPEED CONTROL
The main idea of frequencybased design is to
use the Bode plot of the openloop transfer
function to estimate the closedloop response.
Adding a controller to the system changes the
openloop Bode plot, therefore changing the
closedloop response.
J=0.01;
b=0.1;
K=0.01;
R=1;
L=0.5;
num=K;
den=[(J*L) ((J*R)+(L*b)) ((b*R)+K^2)];
bode(num,den)
APPLICATIONS
Robotics
DC motors are inexpensive, small, and
powerful motors that are widely used.
DC motors are widely used in robotics
because of their small size and high
energy output. They are excellent for
powering the drive wheels of a mobile
robot as well as powering other
mechanical assemblies.
MINIATURE DC MOTORS IDEAL FOR
COMPACT CONSUMER ELECTRONICS
THANK YOU
COMPENSATION
O
The design of a control system is concerned with the arrangement
of the system structure and the selection of a suitable
components and parameters.
O
A compensator is an additional component or circuit that is
inserted into a control system to compensate for a deficient
performance.
O
Types of Compensation
¤
Cascade compensation
¤
Feedback compensation
¤
Output compensation
¤
Input compensation
327
PID CONTROLLERS
O
PID control consists of a proportional plus derivative (PD)
compensator cascaded with a proportional plus integral (PI)
compensator.
O
The purpose of the PD compensator is to improve the transient
response while maintaining the stability.
O
The purpose of the PI compensator is to improve the steady state
accuracy of the system without degrading the stability.
O
Since speed of response, accuracy, and stability are what is
needed for satisfactory response, cascading PD and PI will suffice.
328
THE CHARACTERISTICS OF P, I, AND D CONTROLLERS
NOTE THAT THESE CORRELATIONS MAY NOT BE EXACTLY ACCURATE, BECAUSE KP, KI, AND KD
ARE DEPENDENT OF EACH OTHER. IN FACT, CHANGING ONE OF THESE VARIABLES CAN CHANGE
THE EFFECT OF THE OTHER TWO. FOR THIS REASON, THE TABLE SHOULD ONLY BE USED AS A
REFERENCE WHEN YOU ARE DETERMINING THE VALUES FOR KI, KP AND KD.
Response Rise Time Overshoot Settling Time SS Error
K
P
Decrease Increase
Small Change
Decrease
K
I
Decrease Increase Increase Eliminate
K
D
Small Change
Decrease Decrease
Small Change
329
THE SIMPLEST FORM OF COMPENSATION IS GAIN
COMPENSATION
05 . 0
10
5 . 0 5 . 0 5 . 0 5 . 0
) (
1
0.5; 0.5 
at be will poles loop closed the ratio damping specified the achieve that to shows
geometry Simple . 2 1/ Chose locus; root the Draw 5%; P.O : t requiremen Design
function step a error to zero exhibit will system loop  closed the 1; type is system The
system this of response transient the improve to is goal Our
) 1 (
10
) (
5 . 0 5 . 0
·
+ + −
· · t ·
·
+
·
+ − ·
j j
s G
K j s
s s
s G
j s
p
p
ξ
330
G
c
G
p
R

+
ROOT LOCUS FOR SIMPLE GAIN COMPENSATOR
O
.
331
0.5
1
0.5
Re (s)
Im(s)
LEAD/LAG COMPENSATION
O
Lead/Lag compensation is very similar to PD/PI, or PID control.
O
The lead compensator plays the same role as the PD controller,
reshaping the root locus to improve the transient response.
O
Lag and PI compensation are similar and have the same response:
to improve the steady state accuracy of the closedloop system.
O
Both PID and lead/lag compensation can be used successfully, and
can be combined.
332
LEAD COMPENSATION TECHNIQUES BASED ON THE ROOTLOCUS
APPROACH
O
From the performance specifications, determine the desired location for the
dominant closedloop poles.
O
By drawing the rootlocus plot of the uncompensated system ascertain
whether or not the gain adjustment alone can yield the desired closedloop
poles. If not calculate the angle deficiency. This angle must be contributed by
the lead compensator.
O
If the compensator is required, place the zero of the phase lead network
directly below the desired root location.
O
Determine the pole location so that the total angle at the desired root
location is 180
o
and therefore is in the compensated root locus.
O
Assume the transfer function of the lead compensator.
O
Determine the openloop gain of the compensated system from the
magnitude conditions.
333
LEAD COMPENSATOR USING THE ROOT LOCUS
1 . 8
2
) 25 . 3 ( ) 23 . 2 (
;
) 6 . 3 (
) 1 (
) ( ) (
6 . 3
1
) ( ; 38 ; 142 180 ; 142 90 116 2
1  at location desired the below directly r compensato the of zero the place We
2 1 ˆ as location root dominant desired a choose will We
1 ; 4
4
0.32; be should %; 35 . ; 4
) ( network, a with system this compensate to desire We
axis in the is locus root The : 0 1 ) ( 1 ; ) (
2
1
2
1
o
1 1
2
1
2
1
· ·
+
+
·
+
+
· · − − · − · + × ·
· ·
t ·
· · · ≥ ≤ ≤
+
+
·
· + · + ·
K
s s
s K
s G s GH
s
s
s G
z s
j  r , r
T O P s Ts
p s
z s
s G
jω
s
K
s GH
s
K
s GH
c
c
o
p p
o
n
n
s
c
θ θ φ
ξ ω
ξ ω
ξ
334
1
θ
p
s = p = 3.6
ADDING LEAD COMPENSATION
THE LEAD COMPENSATOR HAS THE SAME PURPOSE AS THE PD
COMPENSATOR: TO IMPROVE THE TRANSIENT RESPONSE OF THE CLOSED
LOOP SYSTEM BY RESHAPING THE ROOT LOCUS. THE LEAD COMPENSATOR
CONSISTS OF A ZERO AND A POLE WITH THE ZERO CLOSER TO THE ORIGIN
OF THE S PLANE THAN THE POLE. THE ZERO RESHAPES A PORTION OF THE
ROOT LOCUS TO ACHIEVE THE DESIRED TRANSIENT RESPONSE. THE POLE IS
PLACED FAR ENOUGH TO THE LEFT THAT IT DOES NOT HAVE MUCH
INFLUENCE OF THE PORTION INFLUENCED BY THE ZERO.
180 ) ( ) (
0 ;
is r compensato for the formular general The 5%; be to P.O Expect ; 2 1/
3. 3  at poles loop  closed the place we tp, desired the achieve To
1.0s 20%; P.O : ions Specificat Design
) 1 (
10
Consider
3 3
− · ∠
〈 〈
+
+
·
·
t ·
≤ ≤
+
·
t − · j s p c
c
c
p
p
s G s G
b a
b s
a) (s K
(s) G
j s
t
s s
G
ξ
335
ROOT LOCUS FOR SIMPLE GAIN COMPENSATOR
O
.
336
Re (s)
Im(s)
Closedloop poles
3
3
18
) 3 ( 8 . 7
) (
8 . 7
3 10
18 1
;
18
) 3 (
) (
18 15 3
3 . 11 tan
3
3 ; 3 . 11 78.7  90 3;  at s Fix
7 . 78 180 ; 180
3 3
o
2 1 2 1
+
+
·
·
,
`
.

+
+ +
·
+
+
·
· + · + · · ·
· − + · − · − − −
+ ·
s
s
s G
s
s s s
Kc
s
s K
s G
b
c
j s
c
c
o
o
o o o
β
θ θ β α θ θ β α
337
b
a
1 0
s
s+b
s+a s+1
θ
1
θ
2 α β
ADDING A LAG CONTROLLER
O
A firstorder lag compensator can be designed using the root locus. A lag
compensator in root locus form is given by
O
where the magnitude of z
o
is greater than the magnitude of p
o
. A phase
lag compensator tends to shift the root locus to the right, which is
undesirable. For this reason, the pole and zero of a lag compensator must
be placed close together (usually near the origin) so they do not
appreciably change the transient response or stability characteristics of
the system.
o
o
p s
z s
s G
−
−
· ) (
338
HOW DOES THE LAG CONTROLLER SHIFT THE ROOT
LOCUS TO THE RIGHT?
O
Recall finding the asymptotes of the root locus that lead to the zeros at
infinity, the equation to determine the intersection of the asymptotes
along the real axis is:
O
When a lag compensator is added to a system, the value of this
intersection will be a smaller negative number than it was before. The net
number of zeros and poles will be the same (one zero and one pole are
added), but the added pole is a smaller negative number than the added
zero. Thus, the result of a lag compensator is that the asymptotes'
intersection is moved closer to the right half plane, and the entire root
locus will be shifted to the right.
zeros poles
zeros poles
−
−
·
∑ ∑
α
339
CONTROL MODES
THERE ARE MANY WAYS BY WHICH A CONTROL UNIT CAN REACT TO
AN ERROR
AND SUPPLY AN OUTPUT FOR CORRECTING ELEMENTS.
O
The twostep mode: The controller is just a switch which is
activated by the error signal and supplies just an onoff correcting
signal. Example of such mode is the bimetallic thermostat.
O
The proportional mode (P): This produces a control action that is
proportional to the error. The correcting signal thus becomes bigger
the bigger the error. Therefore, the error is reduced the amount of
correction is reduced and the correcting process slows down. A
summing operational amplifier with an inverter can be used as a
proportional controller.
O
The derivative mode: This produces a control action that is
proportional to the rate at which the error is changing. When there
is a sudden change in the error signal the controller gives a large
correcting signal. When there is a gradual change only a small
correcting signal is produced. An operational amplifier connected as
a differentiator circuit followed by another operational amplifier
connected as an inverter make an electronic derivative controller
circuit.
340
O
The integral mode (I): This produces a control action that is
proportional to the integral of the error with time. Therefore, a
constant error signal will produce an increasing correcting signal.
The correction continues to increase as long as the error persists.
O
Combination of modes: Proportional plus derivative modes
(PD), proportional plus integral modes (PI), proportional plus
integral plus derivative modes (PID). The term threeterm
controller is used for PID control.
O
The controller may achieve these modes by means of pneumatic
circuits, analog electronics involving operational amplifiers or by
the programming of a microprocessor or computer.
341
DC MOTOR SPEED MODELING
THE DC MOTOR HAS BEEN THE WORKHORSE IN INDUSTRY FOR MANY
REASONS INCLUDING GOOD TORQUE SPEED CHARACTERISTICS. IT IS A
COMMON ACTUATOR IN CONTROL SYSTEMS. IT DIRECTLY PROVIDES
ROTARY MOTION AND, COUPLED WITH WHEELS OR DRUMS AND CABLES,
CAN PROVIDE TRANSITIONAL MOTION. THE ELECTRIC CIRCUIT OF THE
ARMATURE AND THE FREE BODY DIAGRAM OF THE ROTOR ARE SHOWN IN
THE FOLLOWING FIGURE.
WE DEVELOP HERE THE TRANSFER FUNCTION OF A SEPARATELY EXCITED
ARMATURE CONTROLLED DC MOTOR.
θ
`
k
342
Motor
Generator
Mechanical
energy
(T, ω )
V
A
R
A
I
A
V
F
R
F
L
F
I
F
L
J motor & load
T
ω
Field circuit Armature circuit
PHYSICAL PARAMETERS
O
Electrical Resistance R= 1 Ω
O
Electrical Inductance L = 0.5 H
O
Input Voltage V
O
Electromotive Force Constant K = 0.01 nm/A
O
Moment of Inertia of the Rotor J = 0.01 kg.m
2
/s
2
O
Damping Ratio of the Mechanical System b = 0.1 Nms
O
Position of the Shaft θ
O
The rotor and shaft are assumed to be rigid
O
The motor torque T is related to the armature current by a constant Kt
O
The back emf, e, is related to the rotational velocity
θ
θ θ
θ
`
` ` `
`
K V Ri
dt
di
L
Ki b J
K e
i K T
e
t
− · +
· +
·
·
343
SPEED CONTROL
O
Speed Control by Varying Circuit Resistance: The operating speed can only
be adjusted downwards by varying the external resistance, R
ext
O
Speed Control by Varying Excitation Flux:
O
Speed Control by Varying Applied Voltage: Wide range of control 25:1; fast
acceleration of high inertia loads.
O
Electronic Control.
( )
rad/s
2 2
φ φ φ
ω
a
ext a
a a
a a a
m
k
R R
k
Va
k
I R V +
− ·
−
·
1
2
2
1
φ
φ
ω
ω
·
m
m
344
TRANSFER FUNCTION
345
2
) )( (
) ( ) ( ) (
) ( ) ( ) (
K R Ls b Js
K
V
s Ks V s I R Ls
s KI s b Js s
+ + +
·
− · +
· +
θ
θ
θ
Controller Plant
R
θ
`
u
DATA MEASUREMENT
O
Once we have identified the transfer function of the system we
may proceed to the final two phases of the design cycle, the
design of a suitable controller and the implementation of the
controller on the actual system. In the case of speed control of the
DC motor, the control will prove to be quite easy.
O
An important point to be highlighted here is that if we have a good
model of the plant to be controlled, and we already have identified
the parameters of the model, then the design of the controller is
easy.
346
Computer
D/A Converter Power Amplifier
DC Motor
Oscilloscope
Armature voltage
Tachometer
DESIGN NEEDS
O
The uncompensated motor may only rotate at 0.1 rad/sec with an input voltage of 1
V. Since the most basic requirement of a motor is that it should rotate at the desired
speed, the steadystate error of the motor speed should be less than 1%.
O
The other performance requirement is that the motor must accelerate to its steady
state speed as soon as it turns on. In this case, we want it to have a settling time of
2 seconds for example. Since a speed faster than the reference may damage the
equipment, we want to have an overshoot of less than 5%. If we simulate the
reference input (r) by a unit step input, then the motor speed output should have:
O
Settling time less than 2 seconds
O
Overshoot less than 5%
O
Steadystate error less than 1%
Use the MATLAB to represent the open loop response
347
PID DESIGN TECHNIQUE FOR DC MOTOR SPEED
CONTROL
O
Design a PID controller and add it into the system.
O
Recall that the transfer function for a PID controller is:
O
See how the PID controller works in a closedloop system using the
previous Figure. The variable (e) represents the tracking error, the
difference between the desired input value (R) and the actual output
(y). This error signal (e) will be sent to the PID controller, and the
controller computes both the derivative and the integral of this error
signal.
O
The signal (u) just past the controller is equal to the proportional gain
(Kp) times the magnitude of the error plus the integral gain (K
i
) times
the integral of the error plus the derivative gain (K
d
) times the
derivative of the error.
s
K s K s K
s K
s
K
K
I P D
D
I
P
+ +
· + +
2
348
THE PID ADJUSTMENT STEPS
O
Use a proportional controller with a certain gain. A code should be
added to the end of mfile.
O
Determine the closedloop transfer function.
O
See how the step response looks like.
O
You should get certain plot.
O
From the plot you may see that both the steadystate error and the
overshoot are too large.
O
Recall from the PID characteristics that adding an integral term will
eliminate the steadystate error and a derivative term will reduce the
overshoot. Let us try a PID controller with small K
i
and K
d
.
O
The settling time is too long. Let us increase K
i
to reduce the settling
time.
O
Large K
i
will worsen the transient response (big overshoot). Let us
increase K
d
to reduce the overshoot.
O
See the plot now and see if design requirements will be satisfied.
349
 LALIT SETHIA
¾ BIOTECHNOLOGY
Controls Systems and
Biotechnology
INTRODUCTION
The synthesis of biotechnology, life sciences and
control engineering is receiving increasing interest.
Though relevant synthesis examples have been
reported for almost forty years (e.g. the dynamical
modeling of genetic feedback control systems
[1968] and the control of enzyme activity [1970]),
the application of control and systems theory to
biology (now labelled systems biology) has received
increasing impetus in the past five years. A number
of strands of activity are outlined.
Advances in genetic engineering have generated great
interest in biotechnology systems. Modern
fermentation processes have a more scientific basis
and can be optimized more quickly by utilizing
instrumentation and control technology that permits
increasing yield and product quality.
Modern computer systems have made fermentation
processes easier and more accurate by performing
tasks such as online analysis, statistical process
control and supervisory control, while microprocessor
based distributed process controllers perform direct
digital control, batch and sequencing control.
REQUIREMENTS BY BIOTECHIES
O
At the process level, the infrastructure of
biotechnology control systems is based on the
technology used by the process instrumentation,
which includes sensors, transmitters, transducers,
process controllers, online analyzers and final
control elements.
O
At the facility production level, the infrastructure of
biotechnology control systems is based on information
systems communications technology applied to the
integration of the facility's manufacturing requirements.
O
Biotechnology control systems require a single, affordable,
digital communications infrastructure capable of
supporting plant operations.
O
Biotechnology automation systems utilizing Computer Integrated
Manufacturing (CIM) technology include Manufacturing Execution
Systems (MES) and Material Resource Planning (MRP II) Systems,
which are directed to optimize plant manufacturing operations.
O
The proper design of the plantwide distributed control system
architecture is measured in terms of safety, reliability, security,
compliance with current Good Manufacturing Practices (cGMP's),
ease of validation, process flexibility, process scalability, system
flexibility, performance and expandability.
CONTROL NETWORK
COMMUNICATION
O
Digital control network communication can be performed through shared
media or switchbased communication technology. Digital sharedmedia
communication technology allows only one device to transmit information
at any given time. Although this technology has limitations, it supports
most of the present biotechnology requirements. There are two widely
used digital sharedmedia communication methods that define how
network devices communicate with each other: peertopeer and
client/server systems. A common peertopeer system is a tokenring
passing scheme (IEEE 802.5), where only the device that has the token
can transmit or request information. The figure in the next slide shows an
example of a token ring system architecture for a biotechnology facility.
Once the device completes the transmission, the token is passed to
another device in the network in a predetermined sequence.
O
At the heart of the control and information network management system
is a set of software applications that provide network status monitoring,
configuration control, fault detection, system security, user access
capability and long term data storage. In biotechnology control and
information systems, network management must be performed from one
dedicated platform.
COMPUTER INTEGRATED
MANUFACTURING
O
A major obligation of biotechnology facility operations personnel is to
trace raw materials and intermediate products through manufacturing
processes that include manual and semiautomated procedures.
Normally, intermediate products are processed in different
manufacturing areas. For example, media preparation, fermentation,
purification, fill/finish and packaging operations would be performed
in different facility locations. Also, operations personnel must
correlate laboratory sample information to the products sampled to
ensure that quality control is maintained. Production planning must
also be performed to coordinate the facility resources in equipment
and personnel with production schedules.
O
The DCS provides limited capability for facility wide planning,
scheduling, and tracking raw materials and products. Normally,
process control systems do not include enough information
management resources to plan, coordinate and document the
overall manufacturing environment in which they execute.
DISTRIBUTED CONTROL SYSTEM FEATURES
O
Distributed control systems can minimize instrumentation wiring costs by
locating the data acquisition and controlling elements near the process,
and by transmitting the process information to plant operations
management via redundant communication channels. This concept has
great appeal to industries that have automated processes installed
throughout the facility, because wiring cost savings are substantial. In
addition to the potential wiring cost savings, DCSs use redundant
microprocessors to perform direct digital control tasks enhancing the
system reliability and availability. In a DCS, redundancy of critical control
elements can be achieved at a lower cost than adding redundancy in a
mainframe computer system.
O
DCSs are an excellent vehicle for automating biotechnology plants
because of their inherent reliability, and because of the large number of
plant operation management features offered at a reasonable cost.
However, in biotechnology facilities with biocontainment level three (BL3)
requirements, and serious space limitations, it may not be possible to
distribute the control system remote I/O equipment throughout locations
adjacent to the manufacturing suites.
BIBLIOGRAPHY
1. DEVELOPMENT OF BIOTECHNOLOGY
CONTROL SYSTEMSFRANK J. ROMEU
2. PROGRESS IN ARTIFICIAL CONTROL SYSTEM
IN GENE EXPRESSIONZHANG YONG, HAO
NANMING & LIU QIANG.
3. BIOTECHNOLOGY, LIFE SCIENCES AND
CONTROL ENGINEERING – SOME SYNTHESIS
ISSUES.