ECE 3151 - Spring 2010

Performance of Feedback Control

Systems

Prof. Marian S. Stachowicz

Laboratory for Intelligent Systems
ECE Department, University of Minnesota Duluth

February 25 – March 2, 2010

 Outline

• Introduction
• Test Input Signals
• Performance of a second-order system
• Effects of a Third Pole and a Zero on the Second-
Order System Response
• Estimation of the Damping Ratio
• The s-plane Root Location and the Transient
Response

Control Systems 2
 References for reading

1. R.C. Dorf and R.H. Bishop, Modern Control Systems,

11th Edition, Prentice Hall, 2008,
Chapter 5.1 - 5.12

2. J.J. DiStefano, A. R. Stubberud, I. J. Williams,

Feeedback and Control Systems, Schaum's Outline
Series, McGraw-Hill, Inc., 1990
Chapter 9

Control Systems 3
 Test Input Signal

• Since the actual input signal of the system

is usually unknown, a standard test input
signal is normally chosen. Commonly
used test signals include step input, ramp
input, and the parabolic input.

Control Systems 4
General form of the standard test
signals

r(t) = tn

R(s) = n!/sn+1

Control Systems 5
 Test signals r(t) = A tn

n=0 n=1 n=2

r(t) = A r(t) = At r(t) = At2
R(s) = A/s R(s) = A/s2 R(s) = 2A/s3

Control Systems 6
 Table 5.1 Test Signal Inputs

Test Signal r(t) R(s)

Step r(t) = A, t > 0 R(s) = A/s

position = 0, t < 0
Ramp r(t) = At, t > 0 R(s) = A/s2
velocity = 0, t < 0
Parabolic r(t) = At2, t > 0 R(s) = 2A/s3
acceleration = 0, t < 0
Control Systems 7
Test inputs vary with target type

step

ramp
parabola

Control Systems 8
 Steady-state error

• Is a difference between input and the

output for a prescribed test input as

t 

Control Systems 9
 Application to stable systems

• Unstable systems represent loss of

control in the steady state and are
not acceptable for use at all.

Control Systems 10
Steady-state error:
a) step input, b) ramp input

Control Systems 11
 Time response of systems

c(t) = ct(t) + css(t)

The time response of a control system is divided

into two parts:
• ct(t) - transient response

• css(t) - steady state response

Control Systems 12
 Transient response

• All real control systems exhibit transient

phenomena to some extend before steady
state is reached.

lim ct(t) = 0 for t 

Control Systems 13
 Steady-state response

• The response that exists for a long time

following any input signal initiation.

Control Systems 14
Poles and zeros of a first order system

Css(t) Ct(t)
Control Systems 15
 Poles and zeros

1. A pole of the input function generates the form of the

forced response ( that is the pole at the origin generated
a step function at the output).
2. A pole of the transfer function generate the form of the
exponential response
3. The zeros and poles generate the amplitudes for both the
transit and steady state responses ( see A, B in partial
fraction extension)

Control Systems 16
Effect of a real-axis pole upon transient
response

A pole on the real axis generate an exponential response

of the form Exp[-t] where - is the pole location on real axis.
The farther to the left a pole is on the negative real axis,
the faster the exponential transit response will decay to zero.
Control Systems 17
Evaluating response using poles

K1 K 2 K3 K4
C(s)    
s s 2 s 4 s 5
Css(t) Ct(t)
2t 4 t 5t
c(t)  K1  K 2e  K 3e  K 4e
Control Systems 18
 First order system

a at
C(s)  R(s) G(s)  c(t)  1 e
s(s  a)

Control Systems 19
First-order system response to a unit step

Transient response specification:

1. Time-constant, 1/a
2. Rise time, Tr
3. Settling time, Ts
Control Systems 20
 Transient response specification
for a first-order system

1. Time-constant, 1/a
Can be described as the time for (1 - Exp[- a t])
to rise to 63 % of initial value.

1. Rise time, Tr = 2.2/a

The time for the waveform to go from 0.1 to 0.9
of its final value.

3. Settling time, Ts = 4/a

The time for response to reach, and stay within,
2% of its final value
Control Systems 21
Transfer function via laboratory testing

K
G(s) 
(s  a)
K K K
C(s)   a a
s(s  a) s (s  a)


Control Systems 22
 Identify K and a from testing

The time for amplitude to reach 63% of its final value:

63 x 0.72 = 0.45, or about 0.13 sec , a = 1/0.13 = 7.7

From equation, we see that the forced response reaches

a steady-state value of K/a =0.72 .
K= 0.72 x 7.7= 5.54

G(s) = 5.54/(s+7.7) .

Control Systems 23
 Exercise

A system has a transfer function

G(s)= 50/(s+50).

Find the transit response specifications

such as Tc, Tr, Ts.

Control Systems 24
Control Systems 25
 Steady-state response

• If the steady-state response of the output does not

agree with the steady-state of the input exactly, the
system is said to have a steady-state error.

• It is a measure of system accuracy when a specific type

of input is applied to a control system.

Control Systems 26
 Y(s) = R(s) G(s)

Control Systems 27
 Steady-state error

T(s) = 9/(s + 10) Y(s) = 9/s(s+10)

y(t) = 0.9(1- e-10t)

y(∞) = 0.9
E(s) = R(s) - Y(s)
ess = lim s 0 s E(s) = 0.1

Control Systems 28
 R(s) R(s)
E(s)  E(s) 
1 K K
1
1 s
e(t) 
1 K e(t)  eK t
 Control Systems 29

Control Systems 30
Control Systems 31
Performance of a second-order system

Control Systems 32
Numerical example of the second-
order system

Control Systems 33
 Overdamped

9 9
C(s)  2 
s(s  9s  9) s(s  7.854)(s  1.146)

c(t)  1 0.171e7.854 t  1.17e1.146t


Control Systems 34
 Underdamped

9 9
C(s)  2 
s(s  2s  9) s(s  1 j 8)(s  1 j 8)

t 8
c(t)  1 e (cos 8t  sin 8t)
8
 Control Systems 35
 t 8
c(t)  1 e (cos 8t  sin 8t)
8
Control Systems 36
 Undamped

9
C(s)  2
s(s  9)

c(t)  1 cos 3t
Control Systems 37


 Critically damped

9 9
C(s)  2 
s(s  6s  9) s(s  3) 2
3t 3t
c(t)  1 3te e
 Control Systems 38
Step response for second order system
damping cases

Control Systems 39
Control Systems 40
fig_04_11
 Summary

• Overdamped
Poles: Two real at - 1, - 2
• Underdamped
Poles: Two complex at - d + jd, - d - jd
• Undamped
Poles: Two imaginary at + j1, - j1
• Critically damped
Poles: Two real at - 1,

Control Systems 41
Performance of a second-order system

Control Systems 42
 Response to unit step input
G(s)
Y (s)  R(s)
1 G(s)

n 2
Y (s)  2 R( s)
s  2 n s   n
2


n 2
Y (s) 
s ( s 2  2 n s   n )
2

  1 2
1
y (t )  1  e  nt sin( n  t   )
   cos 1 
Control Systems 43
• Natural frequency n - the frequency of natural
oscillation that would occur for two complex
poles if the damping were equal to zero

• Damping ratio  - a measure of damping for

second-order characteristic equation

Control Systems 44
 Characteristic equation
s  2 n s   n  0
2 2

s1   n   n  2 1
s2   n   n  2 1



Control Systems 45
Finding n and  for a second-order
system
36 s  2 n s   n  0
2 2
G(s)  2
s  4.2s  36 s1   n   n  2 1
n 2
G(s)  2 s2   n   n  2 1
s  2 n s   n
2

 n 2  36 
2 n  4.2
n  6
  0.35 Control Systems 46
Second-order responses for 

underdamped

Control Systems 47
Control Systems 48
Transit response

For step input as

a function of 

For step input as

a function of  and nt

Control Systems 49
 Unit impulse response

n2
Y(s)  2 2 R(s)
(s  2 n s   n )

R(s)=1 T(s)=Y(s)

 n  t
y(t)  e sin  n t
n

Control Systems 50
Control Systems 51
 Constant real part

Constant imaginary part

Constant damping ratio

Control Systems 52
 Standard performance measures
4
Ts (s)  4  
 n
Settling time
 Peak time
T p

n 1   2

1 2
Peak response
M pt  1 e

1 2
Percent overshoot
P.O.  100e

Control Systems 53
Control Systems 54
fig_04_14
 Settling time

• The settling time is defined as the time

required for a system to settle within a
certain percentage of the input amplitude.

4
Ts (s)  4  
 n

Control Systems 55

Settling time

4
Ts (s)  4  
 n

Control Systems 56
 Rise time

• The time it takes for a signal to go from

10% of its value to 90% of its final value

2.16  0.60
Tr (s)  0.3    0.8
n

Control Systems 57
Rise time

Control Systems 58
 Peak time

• Peak time is the time required by a signal

to reach its maximum value.


T p
n 1   2

Control Systems 59
 Peak time


T p
n 1   2
Control Systems 60
 Percent overshoot

• Percent Overshoot is defined as:

P.O. = [(Mpt – fv) / fv] * 100%

Mpt = The peak value of the time response

fv = Final value of the response


1 2
P.O.  100e
Control Systems 61
Percent overshoot

Control Systems 62
Percent overshoot and normalized
peak time versus 

Control Systems 63
Finding transient response

25
G(s) 
s(s  5)

Control Systems 64
T(s)  2
25 
T p  0.726 sec
s  5s  25  n 1  2


  s  5s  25
2
1 2
  s  2 n s  
2 2 P.O.  100e  16.3%
n

4
Ts (s)   1.6 sec
 n  25  5  n

2 n  5,   0.5
Control Systems 65
Gain design for transient response

K
G(s) 
s(s  5)

Control Systems 66
 K
T(s)  2
s  5s  K for P.O.  10%


  s  5s  K 1 2
P.O.  100e  10.0%
2

  s2  2 n s   n2

  0.591, K  17.9
n  K
5
2 n  5,  
2 K Control Systems 67
Performance Indices

Elevator

Control Systems 68
Simplified description of a control
system

Control Systems 69
 Elevator input and output

When the fourth floor button is pressed on the first

floor, the elevator rises to the fourth floor with a
speed and floor level accuracy designed for
Control Systems 70
passenger comfort.
Push of the fourth-floor button is an input that
represent a desired output, shown as a step
function. Control Systems 71
 Transient response

Passenger comfort and passenger patience are

dependent upon the transient response.
If this response is too fast, passenger comfort
is sacrificed; if too slow, passenger patience is
sacrificed. Control Systems 72
 Steady-state error

Passenger safety and convenience would be

sacrificed if the elevator is not properly level.
Control Systems 73
 Performance Indices

• A performance index is a quantitative

measure of the performance of a
system and is chosen so that
emphasis is given to the important
system specifications.

Control Systems 74
Response of the system

Control Systems 75
ISE - Integral of Square of Error

T
I1   e (t)dt
2


Control Systems 76
 The Integral Squared Error

T
I1   e (t)dt
2

Control Systems 77
IAE - Integral of the Absolute
Magnitude of the Error

T
I2   e(t) dt
0


Control Systems 78
ITAE - Integral of Time Multiplied by
Absolute Error

T
I3   t e(t) dt
0


Control Systems 79
ITSE - Integral of Time Multiplied by
Squared Error

T
I4   te (t)dt
2


Control Systems 80
 General form of the performance
integral

T
I  f [e(t),r(t),c((t),t]dt
0

 Control Systems 81
 Section 5.9
T
T
ISE   e (t )dt 2
IAE   | e(t ) | dt
0 0

T T

ITAE   t | e(t ) | dt ITSE   te 2 (t )dt

0 0

T
I   f (e(t ), r (t ), y (t ), t )dt
0

Control Systems 82
 Performance criteria
1
T(s)  2
s  2s  1

Control Systems 83
Control Systems 84
 Optimum system

• A control system is optimum when the

elected performance index is minimized.

• The optimum value of the parameters

depends directly upon the definition of
optimum, that is, the performance index.

Control Systems 85
• The coefficients that will minimize the
ITAE performance criterion for a step
input and a ramp input have been
determined for the general closed-loop
transfer function.

Control Systems 86
 General closed-loop T(s)

Y (s) b0
T(s)   n
R(s) s  bn1sn1  ... b1s  b0

The T(s) has n poles and no zeros.



This T(s) has a steady-state error equal zero for a step input.

Control Systems 87
 Table 5.6 The Optimum
Coefficients of T(s) Based on the
ITAE Criterion for a Step Input
• s + ωn
• s2 + 1.4ωns + ωn2
• s3 + 1.75ωn s2 + 2.15ωn2s + ωn3
• s4 + 2.1 ωn s3 + 3.4 ωn2s2 + 2.7 ωn3s + ωn4
• s5 + 2.8 ωn s4 + 5.0 ωn2s3 + 5.5 ωn3s2 + 3.4ωn4s +ωn5
• s6 + 3.25ωns5 + 6.60 ωn2 s4 + 8.60ωn3s3 + 7.45 ωn4s2 + 3.95ωn5s
+ωn6

Control Systems 88
Step response for optimum coefficients

Control Systems 89
Control Systems 90
Table 5.7 The Optimum Coefficients
of T(s) Based on the ITAE Criterion
for a Ramp Input
• s2 + 3.2ωns + ωn2
• s3 + 1.75ωn s2 + 3.25ωn2s + ωn3
• s4 + 2.41 ωn s3 + 4.93 ωn2s2 + 5.14 ωn3s + ωn4
• s5 + 2.19 ωn s4 + 6.50 ωn2s3 + 6.30 ωn3s2 + 5.24ωn4s +ωn5

Control Systems 91
 Y (s) b1s  b0
T(s)   n
R(s) s  bn1s  ... b1s  b0
n1

T(s) has a steady-state error equal to zero for a

ramp input.
T(s) has two or more pure integrations as
required to provide zero steady-state error.

Control Systems 92
A: Simplification of linear system

K
G(s) 
s(s  2)(s  30)

We can neglect the impact of the pole at s = - 30 ,

however we must retain the steady-state response
and
reduce the system to

K /30
G(s) 
s(s  2)

Control Systems 93

Impulse response

Control Systems 94
Control Systems 95
 B: Simplification of linear systems

am sm  am1sm1    a1s  1
H(s)  K , mn
 bn s  an1s    b1s  1
n n1

c p s p    c1s  1
L(s)  K , pgn
 dg s    d1s  1
g



Control Systems 96
 dk
M (k )
(s)  k M (s)
ds

k
d
( k ) ( s )  k ( s )
ds

(1) k  q M k (0) M ( 2 q  k ) (0)

2q
M 2q 
k 0 k!(2q  k )!

M 2q   2q q  1,2...

Control Systems 97
 Simplified model
6 1
H (s)  3 
s  6 s  11s  6 1  (11 / 6) s  s 2  (1 / 6) s 3
2

6
L(s) 
1 d1s  d2 s2

M ( s)  1  d1 s  d 2 s 2

( s )  1  (11 / 6) s  s 2  (1 / 6) s 3

M (0)
( s)  1  d1 s  d 2 s 2
M (0) (0)  1
Control Systems 98
 Example 5.9

d
M ( k ) ( s)  (1  d1 s  d 2 s 2 )  d1  2d 2 s
ds

M (0)
( 0)  1 ( 0 ) (0)  1

M (1)
( 0)  d 1  (0)  11 / 6
(1)

M ( 2 ) ( 0)  2 d 2  ( 0)  2
( 2)

M ( 3 ) ( 0)  0  (0)  1
( 3)

Control Systems 99
 Example 5.9
M 0 (0) M 2 (0) M 1 (0) M 1 (0) M 2 (0) M 0 (0)
M 2  (1)   (1)
2 1 2

2 2
M 2   d 2  d1  d 2  2d 2  d1

0 (0)2 (0) 1 (0)1 (0) 2 (0)0 (0) 49

 2  (1)   (1) 
2 1 2 36

Control Systems 100

 Example 5.9

7
 2d 2  d 1
2

49 d2 
36 18

1 1.60
L( s )  
1  1.165s  0.625s 2
1.60  2.584 s  s 2

Control Systems 101

 Example 5.9

6
H (s)  3
s  6 s 2  11s  6

1.60
L(s) 
1.60  2.584s  s2

Control Systems 102

Control Systems 103
Impulse response

Control Systems 104

Control Systems 105
Dominant poles of transfer function

• It has been recognized in practice and in

the literature that if the magnitude of the
real part of a pole is at least 5 to 10 times
of a dominant pole or pair of complex
dominant poles, than the pole may be
regarded as insignificant insofar as the
transient response is concerned.

Control Systems 106

Thank You.

Control Systems 107