06-01-2024

Performance of Feedback Control Systems

Time response of systems

 In the design of feedback control system our aim is ensure that the
feedback control system meets certain performance specifications. c(t) = ct(t) + css(t)
 Measurable performance specifications include:
. Time domain specifications
The time response of a control system is divided
. Transient-response specifications:
into two parts:
-percentage overshoot
 ct(t) - transient response
-settling time
-peak time
-rise time  css(t) - steady state response
. Steady-state response specifications:
-steady-state errors
. Frequency domain specifications

Transient response Steady-state response

 The response that exists for a long time
 All real control systems exhibit transient following any input signal initiation.
phenomena to some extend before steady
state is reached.

lim ct(t) = 0 for t 

1
 06-01-2024

Test signals r(t) = A tn First order system

n=0 n=1 n=2

r(t) = A r(t) = At r(t) = At2 a
R(s) = A/s R(s) = A/s2 R(s) = 2A/s3 C(s)  R(s) G(s)  c(t)  1 eat
s(s  a)

First-order system response to a unit step

Performance of a Second-Order System

Transient response specification:

1. Time-constant, 1/a
2. Rise time, Tr
3. Settling time, Ts

2
 06-01-2024

n2
Relationship between the characteristic-equation A second-order system: G ( s) 
s  2n s  n2
2

roots and the step response

n2   1: j
A second-order system: G ( s) 
s 2  2n s  n2 s1,2  n  n  2  1 overdamped
0
Its characteristic equation: D( s)  s 2  2n s  n2  0
j
The value of  determines the location of the roots of D ( s )  0.   1: s1,2  n
0 critically damped
  1: s1,2  n  n  2  1 overdamped
0    1: j
  1: s1,2  n critically damped
s1,2  n  jn 1   2
0    1: s1,2  n  jn 1   2 underdamped 0 underdamped

  0 : s1,2   jn j
undamped
  0 : s1,2   jn
0
undamped

Overdamped Underdamped

9 9 9 9
C(s)   C(s)  
s(s2  9s  9) s(s  7.854)(s  1.146) s(s  2s  9) s(s  1 j 8)(s  1 j 8)
2

c(t)  1 0.171e7.854t  1.17e1.146t 8

c(t)  1 et (cos 8t  sin 8t)
8

3
 06-01-2024

Undamped Critically damped

9 9
9 C(s)  
C(s)  2 s(s  6s  9) s(s  3) 2
2

s(s  9)

c(t)  1 cos 3t c(t)  1 3te3t  e3t

Step response for second order system

damping cases Finding n and  for a second-order system

36 s 2  2 n s   n  0
2
G(s) 
s  4.2s  36
2
s1   n   n  2  1
n 2
G(s)  s2   n   n  2  1
s  2 n s   n
2 2

 n 2  36
2 n  4.2
n  6
  0.35

4
 06-01-2024

Time-Domain Specifications Standard performance measures

 Rise time: is the time required for the step response to go from 0
to 100% of its final value (usually unity).
 Peak time: is the time required for the step response to reach the 4
first peak of the overshoot.
Ts (s)  4  
 n
Settling time
 Overshoot: is the difference between the maximum value of the
 Peak time
step response and its steady-state value (usually unity). T p
 Settling time: is the time required for the step response to reach n 1   2
and remain within a certain percentage (usually 2%) of its final value 
(usually unity).
1 2
Peak response
M pt e

1 2
Percent overshoot
P.O.  100e

Example Example
 Consider the system shown in following figure, where Rise Time Peak Time
damping ratio is 0.6 and natural undamped frequency is 5
rad/sec. Obtain the rise time tr, peak time tp, maximum   
tr  tp 
overshoot Mp, and settling time 2% and 5% criterion ts when
d d
the system is subjected to a unit-step input.

Settling Time (2%) Maximum Overshoot

4 
t s  4T  
 n Mp e
1 2
 100
3
t s  3T 
 n
Settling Time (4%)

5
 06-01-2024

Example Example
Rise Time Peak Time
Settling Time (2%)
   4
tr  tp  ts 
d d  n
4
3.141   3.141 ts   1.33s
tr   tp   0.785s 0 .6  5
n 1   2 4
Settling Time (4%)

n 1   2 3
  tan1( )  0.93 rad ts 
n  n
3.141  0.93 3
tr   0.55s
5 1  0. 6 2
ts   1s
0.6  5

Example Example
Maximum Overshoot  For the system shown in Figure, determine the values of gain K and

velocity-feedback constant Kh so that the maximum overshoot in
 the unit-step response is 0.2 and the peak time is 1 sec. With these
1 2
Mp e  100 values of K and Kh, obtain the rise time and settling time. Assume
that J=1 kg-m2 and B=1 N-m/rad/sec.

3.1410.6

1 0.6 2
Mp e  100

M p  0.095  100

M p  9.5%

6
 06-01-2024

Example Example

Since J  1 kgm 2 and B  1 Nm/rad/sec

C( s ) K

R( s ) s 2  (1  KK h )s  K
• Comparing above T.F with general 2nd order T.F

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

(1  KK h )
n  K  
2 K

Example Example
(1  KK h )  n  3.96
n  K  
2 K
• Maximum overshoot is 0.2. • The peak time is 1 sec

 (1  KK h )
n  K  
tp  2 K
d
3.141 3.53  K 0.456  2 12.5  (1  12.5 K h )
1
2

 n 1   3.532  K K h  0.178
)  ln 0.2 
1 2
ln( e
3.141
n 
K  12.5
1  0.456 2

 n  3.53

7
 06-01-2024

Example
n  3.96 Error and steady-state error
Open-loop control system Closed-loop control system
R(s) E(s) Y (s)
4 R(s) Y (s) G(s)
  ts 
G(s)
r(t ) e(t ) y(t )
tr  r(t) y(t )
n 1   2  n
Error: e(t )  r (t )  y (t )
t r  0.65s t s  2.48s
Steady-state error: ess  lim e(t )
t 
Utilizing the final value theorem: lim f (t )  lim sF ( s )
t  s 0
3
ts  ess  lim e(t )  lim sE ( s )
 n t  s 0

Assuming r(t)=1(t) is a unit-step input, according to the above

t s  1.86s definition, could you calculate the steady-state error of the open-
loop and closed-loop control systems?

System Type System Type

 System Type: The value of n in the denominator or, the number of  Example:
pure integrations in the forward path. i. K s  2
G s   Type
s  1s  3
0
ii. K 0.5s  1
G s   Type
ss  12s  1s 2  s  1
1
iii. K 2 s  1
G s   Type
s3
3
 Therefore,
i. If n = 0, system is Type 0
ii. If n = 1, system is Type 1
iii. If n = 2, system is Type 2

8
 06-01-2024

Steady State Error Steady State Error

 Example: Calculate the error constants and determine ess for a  Example: Calculate the error constants and determine ess for a
unit step, ramp and parabolic functions response of the following unit step, ramp and parabolic functions response of the following
system. system.
1 5s  1
G s   ; H s  
s s  12 s5  For step input,
 sR s  
K p  limG( s ) H ( s ) ess  lim 
s  0 1  G s H s  
s 0  
 5s  1   
 lim
s 0 s s  12s  5   s1 s 
 
   lim  
s 0
1  5s  1 

 ss  12s  5 
 1 
 lim  0
s 0 1   
 

Steady State Error Steady State Error

 Example: Calculate the error constants and determine ess for a  Example: Calculate the error constants and determine ess for a
unit step, ramp and parabolic functions response of the following unit step, ramp and parabolic functions response of the following
system. system.

 For ramp input,  For parabolic input,

sR s  
K v  limsG s H (s )
s 0 ess  lim 
 sR s  
s  0 1  G s H  s  
 
K a  lim s 2G s H s  ess  lim 

s  0 1  G s H s  
 
s 0  
 5s  1   5s s  1 
 lim     lim   
s 0 s  12 s  5 
    
s 1 s2  s 0 s  12 s  5 
 

 lim 
 
s 1 s3 

 lim   5s  1

1 s0
1  5s  1  0 s0
1  
12  ss  12s  5   s s  12s  5 
 12 

9
 06-01-2024

Steady State Error

 Example: Find the steady-state errors for inputs of 5u(t), 5tu(t), and
5t2u(t).

5 5 5
estep ( )   eramp ()  
1  lim G ( s) 21 lim sG ( s)
s 0 s 0

5
e para ()  
lim s 2G (s )
s 0

K p  20, K v  0, K a  0

Proportional Function Integral Function

Controller output is proportional to error e.
The controller output u is proportional to error signal e:
1 du t
u  Kce  e  S 0 e  u  S 0  edt
 dt 0

where  is proportional band • Output of I control is constant only e=0, no steady

state error.
P control has steady state error.
• Reduce system stability. I control is always slower
than that of P control,
• Open loop gain is proportional to S0, increase S0
Inreduce system stability.

10
 06-01-2024

PI Control Derivative Function

Output of controller is proportional to the differential of error
P to improve response time and reject disturbance,
I to eliminate steady state error.
de dy
t 1 1 t  u  S2 or u   S2
u  K c e  S0  edt   e  0 edt  dt dt
0   TI
Prediction: Adjusting the output according to speed of error.

D function must be formed to PD or PID controllers.

Controller takes no action if rate of change very small,
accumulate error.

PD and PID Control

de  de 
u  Kce  S2  K c  e  TD 
dt  dt 
1 de 
  e  TD 
 dt 

t de
u  K c e  S 0  edt  S 2
0 dt
 1 t de  1  1 t de 
 K c  e   edt  TD   e   edt  T 
dt   
D
 TI 0 TI 0 dt 

Properties
1. Steady state, de/dt=0, PD control has steady state error.
2. D function reduces oscillation, increases system stability.
3. Adding D increase open loop gain, increase response speed.
4. Sensitive to disturbance.

11