Control Systems (Control Systems Design)

Control Systems Design

Objectives

Upon completion of this chapter you will be able to:

 Design different compensators to meet the frequency domain specifications of any

Control System.
 Design different controllers and understand their effects on time domain
specifications of any Control System.

Introduction

For any system, the specifications of output can be described in terms of time domain
parameters like t , t , t , t ,M & e and in frequency domain using parameters like
d r p s p ss
M ,  , Gain margin, phase margin, Bandwidth. To meet these specifications, process
r r
parameters are not altered but rather we add additional components that can modify these
characteristics to meet the specifications.
Controllers or compensators are additional components that will be added to existing design
to improve the time domain & frequency domain specifications respectively.

Compensators

Compensator is an electrical component which will be introduced in series with the forward
path transfer function in order to introduce certain phase angle into steady state sinusoidal
response of the system.

This property helps us change the phase of the Transfer Function and hence alter the Phase
margin of the system to impact Stability of the system.

© Kreatryx. All Rights Reserved. 1 www.kreatryx.com

 Control Systems (Control Systems Design)

Lead Compensators

Lead Compensators introduce phase lead or positive phase shift into sinusoidal response of
the system.

 
Pole P 
c
1

, Zero Z 
c
1
a  
Zero is closer to origin than pole, so a  1

The pole-zero plot of the Lead Compensator is shown

in the adjoining figure

 
 1
sZs  
c   a 
Transfer Function Gc  s  
s P 
c
 1
s  
 

P
a  c  1,   0
Z
c

G  s   1   as  1
  
c  a   s  1 

 1   ja  1
G  j     
c  a   j  1 

  tan1  a  tan1 

Since, a  1 , so Ф is always positive and thus it provides lead angle.

Maximum phase angle  

m

Frequency for maximum phase angle = 

m

d
For maximum phase shift 0
d

© Kreatryx. All Rights Reserved. 2 www.kreatryx.com

 Control Systems (Control Systems Design)

a 
 0
1  2a2 2 1  2 2

1 1
  
m
a2  a

 a1  a 1
  sin1    tan1  
m a1 2 a

Bode Plot for this system is shown below:

The gain at high frequencies is higher as compared to low frequencies as observed from
Bode Plot so it acts as a High Pass Filter.

Electrical Network

E s
0 
R
2
 R
 2 
 
1 
R Cs  1
E s R R R    R R C  
i 1  1 2   1 2  
s 1
R  sC  R R  
2 1   1 2  
R 
1 sC

R R
a  1 2 1 ;  R C
R 1
2

© Kreatryx. All Rights Reserved. 3 www.kreatryx.com

 Control Systems (Control Systems Design)

Effect of lead compensator

1. Improves the transient response, makes the system response faster.

2. Increases the margin of stability, that means improve the stability.
3. Increases the bandwidth.
4. The noise level in output increases, due to which SNR decreases S / N  S / N
o i

5. It helps to increase the error constant up to some extent, but doesn’t affect steady state
error much.

Lag Compensator

Lag compensator introduces phase lag or negative phase shift into sinusoidal response of
the system.

 
Pole P 
c
1

, Zero Z  c   a1
Pole is closer to origin than zero & hence a  1 .

The pole-zero plot of lag Compensator is shown

in the adjoining figure.

 
 1
sZ s  
c   a 
Transfer Function Gc  s  
 
s P
c
 1
s  
 

P
a c 1
Z
c

 1   as  1
G  s      
c  a   s  1 

 1   ja  1
G  j     
c  a   j  1 

© Kreatryx. All Rights Reserved. 4 www.kreatryx.com

 Control Systems (Control Systems Design)

  tan1  a  tan1 

Since, a  1,  is always negative and thus it provides a lag angle.

Frequency for maximum phase shift is given by:

1
  = GM of both corner frequencies.
m  a

The maximum phase shift is given by the expression:

 a1  a 1
  sin1    tan1  
m a1 2 a

The Bode Plot for this system is shown below:

The gain at lower frequencies is higher as compared to high frequencies as observed from
Bode Plot so it acts as a Low Pass Filter.

Electric Network

1
E s R  R Cs  1
0  2 sC  2
E s
i R R 
1 2 sC
1
1 2 
R  R Cs  1 
R
a 2 1
 R R
1 2 

© Kreatryx. All Rights Reserved. 5 www.kreatryx.com

 Control Systems (Control Systems Design)

Effect of lag compensator

1. Improves the steady state response, it increase error constant and decreases steady state
error.
2. Decrease bandwidth.
3. Reduces effect of noise which means it increases SNR.
4. Reduce stability margin so makes the system less stable.
5. Does not affect transient response.

Lag – Lead Compensator

It is a compensation of lag & lead compensator which provides lag compensation at lower
frequencies & lead compensation at higher frequencies.

The pole-zero plot for Lag-Lead Compensator is shown below:

The Transfer Function for this system can be written as:

 1   1 
s  s
 b2   a1 

Gc s  


s 
1   1
s 


2   1 

a>1 and b<1

© Kreatryx. All Rights Reserved. 6 www.kreatryx.com

 Control Systems (Control Systems Design)

The Bode Plot for this system is shown below:

The gain at edges is more than the gain at middle of band so it acts as Band Stop Filter.

Electric Network

Remember, Lag-Lead compensator cannot be realized using lag and lead compensators
connected in cascade.

If we interchange the location of poles and zeroes we can realize a Lead-Lag Compensator
which provides Phase Lead at lower frequencies and Phase Lag at higher frequencies.

© Kreatryx. All Rights Reserved. 7 www.kreatryx.com

 Control Systems (Control Systems Design)

Summary of Compensators

Compensator Pole-Zero location Type of filter

Lead High Pass

Lag Low Pass

Lead-Lag Band Pass

Lag-Lead Band Reject

Solved Examples

s4
Problem: The TF of a compensator is .The maximum phase lead frequency and the
s  16
corresponding phase is?

 1  0.25s 
Solution: TF= 4  .
 1  .0.0625s 
1  as
Comparing with
1  s
We get   0.0625 and a=0.25 =>a=4
1 1
m   8
 a 0.0625 4
m  8

 s4  1 m 
   tan  tan1 m ; m  36.870
 s  16  4 16

© Kreatryx. All Rights Reserved. 8 www.kreatryx.com

 Control Systems (Control Systems Design)

Problem: Consider the open loop transfer function of unity feedback system as
21s  97
G(s)  . The maximum phase angle provided by this system is?
(s 33)

 21s 
1
97  97 

Solution: G(s)  
33  s 
 1 
 33 

 1  as 
Comparing with standard transfer function k  
 1  s 

1 21 21  33
We get, = and a  =>a=  7.1443
33 97 97

a  1 i.e. Lead Compensator

 a 1 1  7.1443  1 
Maximum phase lead angle   sin    sin    48.97
1 0

a1  7.1443  1 

Controllers

Controllers are the components that are connected in the forward path to change the steady
state response and transient response of the system.

There are different kind of controllers which have different impact on characteristics of the
system as described below:

© Kreatryx. All Rights Reserved. 9 www.kreatryx.com

 Control Systems (Control Systems Design)

Proportional (P)

In proportional controller, the actuating controller signal is proportional to applied input

error signal.

y t  e t

y t  K e t
p

Y s
K
E s p

Without P controller With P controller

1 k
G s  p
G s 
s  s  10  s  s  10 

2 2
Characteristic equation: s  10s  10  0 Characteristic equation: s  10s  Kp  0

10
  10 rad s and   10
n 2   K rad s and  
n p K
p

Using proportional controller, the transient response can be improved but e

ss cannot be
altered.

Derivative (D)

In D controller the actuating signal is proportional to derivative of input error signal.

d
y t  e t
dt
 d 
y  t   k  e  t  
d  dt 

Y s
 k ds
E s

© Kreatryx. All Rights Reserved. 10 www.kreatryx.com

 Control Systems (Control Systems Design)

Without D Controller With D Controller

1 K ds kd
G s  G s  
s2  s  10  s2  s  10  s  s  10 

e = constant for parabolic i/p.

ss 2
Characteristic equation: s  10s  k d  0
G(s) = unstable
For stability k  0 always
d

G (s) = stable and e   for parabolic

ss
input

Using derivative controller stability can be improved but steady state response gets poorer.

Integral (I)

An integral controller gives an output proportional to the integral of input error signal.

y  t    e  t  dt

y  t   k  e  t  dt
i

Y s K
 i
E s s

Without I Controller With I Controller

1 K
i
G s  G s 
2
s  s  10  s  s  10 

System is stable. Characteristic equation: = s3  10s2  K  0

i

e = constant for ramp i/p It is unstable

ss

e  0 for ramp input.

ss

© Kreatryx. All Rights Reserved. 11 www.kreatryx.com

 Control Systems (Control Systems Design)

Using integral control, steady state response can be improved but stability is reduced.

PI Controller

In PI controller, the actuating output is proportional to input error signal and integral of
input error signal.


y  t   e  t    e  t  dt 
y  t   K e  t   K  e  t  dt
p i

Y s K
K  i
E s p s

Without PI Controller With PI Controller

K
1
G s  K  i
s  s  10  p s
G s 
s  s  10 
System is stable
Characteristic Equation :
:
e = constant for ramp input s3  10s2  kps  K  0
ss i

System is stable if 10k  K  0

p i

e  0 for ramp input

ss

PI controller improves steady state response without affecting stability of the system.

PD Controller

In PD controller, the control action is proportional to error signal as well as its derivatives.

 d 
y t   e t  e t 
 dt 

© Kreatryx. All Rights Reserved. 12 www.kreatryx.com

 Control Systems (Control Systems Design)

d
y t  K e t  K
d dt  
e t
p

Y s
  K  K s 
E s  p d 

Without PD Controller With PD controller

1 K K s
G s  p d
G s 
s2  s  10 
s2  s  10 

It is unstable
3 2
Charact. Equation s  10s  K ds  Kp  0
e = constant for parabolic input
ss
Stable if 10k  k  0
d p

e = constant for parabolic input

ss

PD controller improves stability without affecting steady state response of system.

PID Controller

In PID controller, all 3 control actions are incorporated so it can improve transient & steady
state response as well as stability of the system.

© Kreatryx. All Rights Reserved. 13 www.kreatryx.com