MA3005: Control Theory

School of Mechanical and Aerospace

Engineering

Faculty: Asst Prof Tegoeh Tjahjowidodo

Email: [email protected]
Phone: (+65)6790 4952
Office: N3-02c-68
Week 6 – Transient Response
Part I – Lecture Map
2. Laplace
Transform
3. Modelling

1. Introduction to Systems

6. First-Order 4. Response and

Stability

5. Routh Stability
7. Second-Order
8. Transient
Response

9. Basic PART II. CONTROLLER DESIGN

Control
Actions

Week 6 3
Learning Objectives
At the end of this week, you should be able to:
• Calculate features of second-order system and step response
• Analyse direct current (DC) gain, undamped natural frequency and damping ratio of
second-order system

Week 6 4
Performance Indices
Quantitatively related to:
• Delay time and rise time: how fast can output follow input?
• Percentage overshoot: how oscillatory is the response (indicative of damping)?
• Settling time: how long does response take to reach the final value?

Week 6 5
Performance Indices (Cont’d.)

c(t) Allowable tolerance

Mp 0.05
1 or
0.02
td
0.5

0
tr t

tp
ts

This topic for second-order systems is mainly focused on underdamped cases.

Week 6 6
Transient Response Characteristics – Part 1
Transient Response Characteristics
Second-Order System With 0 < z < 1
Transient response characteristics of control system to unit step input:
• Delay time - td
• Rise time - tr
• Peak time - tp
• Maximum overshoot - Mp
• Settling time - ts
This applies only for second-order system with 0 < z < 1 (underdamped cases).

Week 6 8
Delay Time - td
Delay time is the time required for response to reach half of its steady-state value for the very
first time.
c(t) Allowable tolerance
Mp 0.05
1 or
0.02
td
0.5

0
tr t

tp
ts

Week 6 9
Rise Time - tr
Rise time is the time required for underdamped response
Poles: s1,2  z n  j n 1  z 2
to rise from 0 to 100 percent of steady-state value:
   j d
1 1   d

tr  tan  
d   
     cos 1 z
 
d d

Week 6 10
Angle  and Damping Coefficient z

Poles: s1,2  z n  j n 1  z 2

   j d
jd  jωn 1  ζ 2
n
zn
 cos 
 n
– zn  z  cos 

– jd

Week 6 11
Angle  and Damping Coefficient z (Cont’d.)
Rise time depends on both z and n.
Poles: s1,2  z n  j n 1  z 2
For constant zn, the speed of response can be improved (i.e.    j d
rise time can be made smaller) by increasing the imaginary
part d:
1 1   d 
tr  tan  
d    
     cos 1 z
 
d d

Week 6 12
Peak Time - tp
Peak time is the time required for response to reach first
Poles: s1,2  z n  j n 1  z 2
peak of the overshoot:
     j d
tp  
n 1z 2 d
Peak time tp corresponds to half-period of damped oscillation:
1  2 
tp   
2 d 

Week 6 13
Maximum Overshoot - Mp
In maximum peak value of response, it is common to use
percentage maximum overshoot, which is defined as:
c(t p )  c()
%M p   100%
c ( )
Percentage maximum overshoot is a good indicator of the
system damping (relative stability):
z 
 
1z 2
d
%M p  e 100%  e 100%

Week 6 14
Evaluation of System Damping
Damping can be found from inverse of the equation for system overshoot:
 ln( M p )
z 
 2  ln 2 ( M p )

System damping may be required if:

• System response specifications are given in Mp
• System identification is required from the time response plot

Week 6 15
Settling Time - ts
Settling time is the time required for the response to reach
Poles: s1,2  z n  j n 1  z 2
and stay within a specified allowable tolerance (usually
two or five percent) of its final value.    j d

Settling time is commonly defined to be:

4 4
t s  4T   (Two percent criterion)
 z n
3 3
t s  3T   (Five percent criterion)
 z n

Week 6 16
Remarks
T = 1/zn = time constant of underdamped second-order system.
z is usually determined from the required permissible maximum overshoot. Thus, ts is
determined primarily by the n:
 ln( M p )
z 
 2  ln 2 ( M p )
Duration of transient response can be varied by adjusting n without affecting the maximum
overshoot:
4
ts 
zn

Week 6 17
Response Curve
If the values of td, tr, tp, Mp, and ts are specified, then the shape of response curve is
determined.

c(t) For t > ts, response

remains within this strip.
0.05
𝑀𝑝p
M
or
1
0.02

These points are specified.

0.5

0
t𝑡d𝑑 𝑡t𝑟r 𝑡𝑝tp 𝑡ts𝑠 t

Week 6 18
Remarks
Peak time and maximum overshoot do not apply to an overdamped system.
Desirable transient response often require damping ratio between 0.4 and 0.8:
• z < 0.4  excessive overshoot
• z > 0.8  sluggish response
For rapid transient response, n must be large.
To limit maximum overshoot and to make the settling time small, z should not be too small.

Week 6 19
Transient Response Characteristics – Part 2
Constant ts, tp, and %Mp Lines in s-Plane


z %Mp1 Poles: s1,2  z n  j n 1  z 2
1z 2
%M p  e  100% z1    j d

%Mp2 tp2
d2
z2
d1
tp1 1  2 
tp   
2  d 

2 1
4 4
ts  4T  
 zn
ts2 ts1 (Two percent criterion)

Week 6 21
Constant ts, tp, and %Mp Lines in s-Plane (Cont’d)
(a)
Envelope is the same 3 j
2
1 s-plane

1 Pole
2 motion
3

(b)
Frequency is the same
2 1
j
s-plane

Pole
2 1 motion
(c)
Overshoot is the same j
3
21
s-plane

Pole
21 motion
3

Week 6 22
Examples
Example: B-5-2 (pp. 273)
Consider unit step response of a unity feedback control system, given open loop transfer
function is:
1
G( s ) 
s( s  1 )
Obtain tr, tp, Mp, and ts.
The closed-loop transfer function is:
C ( s) 1 12
 2  2
R( s ) s  s  1 s  2(0.5) 1s  12

Week 6 24
Example: B-5-2 (pp. 273) (Cont’d.)
From closed-loop transfer function:
z = 0.5, n = 1 rad/s  d = 0.866,  = 0.5,  =  / 3
     1.0472
tr    2.42 sec
d 0.866
 
tp    3.63 sec
d 0.866
 (  d )d   ( 0.5 0.866 
    
p   e e 00.163
0.5
0.866)
MMp  ee .163

Week 6 25
Example: B-5-2 (pp. 273) (Cont’d.)

4 4
ts    8 sec
 0.5
(Two percent criterion)

Time (sec)

Week 6 26
Example: B-5-9 (pp. 275)
R(s) 16
16 C(s)
+- 𝑠(𝑠 + 0.8)
s ( s 0.8)

1 + ks

Determine the value of k such that z = 0.5:

C ( s) 16 16
  2
R( s ) s ( s  0.8)  16(1  ks ) s  (0.8  16k ) s  16

From the closed-loop transfer function, n = 4 rad/s.

0.8  16k  2  0.5  4  4
3.2
 k  0.2
16
Similarly, d = 3.46,  =  / 3,  = 2.
Week 6 27
Example: B-5-9 (pp. 275) (Cont’d.)
     3
tr    0.605 sec Unit step response
d 3.46

tp   0.907 sec
d


d  2 3.46 
Mp e e  0.1627

Amplitude
4 4
ts    2 sec
zn 0.5  4
(Two percent criterion)

Time (sec)

Week 6 28
Impulse Response
Impulse Response
Unit impulse response reveals the natural characteristics of the system.
Given r(t) = d (t)  R(s) = 1,
z = 0.1
z = 0.3
C (s)  G(s)
z = 0.5
n 2 z = 0.7
 2 z = 1.0
s  2zn s  n 2
c(t )
n

c(t )
n

Week 6 30
Remarks
Impulse response for various z can be obtained from differentiating the corresponding unit
step response.
For critically damped and overdamped cases, unit impulse response is always positive or zero.
For underdamped case, unit impulse oscillates about zero and takes both positive and
negative values.

Week 6 31
Lesson Summary
At the end of this week, you should be able to:
• Calculate features of second-order system and step response
• Analyse DC gain, undamped natural frequency and damping ratio of second-order system

Week 6 32
Reference
No. Slide No. Reference

Figure B-5-2. An example of response curve. Adapted from Modern Control Engineering (p. 273), by Katsuhiko Ogata, 2009, 5th Edition, New Jersey, NJ:
1. 24-26
Prentice Hall.

Figure B-5-9. An example of response curve. Adapted from Modern Control Engineering (p. 275), by Katsuhiko Ogata, 2009, 5th Edition, New Jersey, NJ:
2. 27-28
Prentice Hall.

Week 6 33