0% found this document useful (0 votes)
215 views

Design of Compensators Using Root Locus Diagram

This document discusses various types of compensators that can be added to control systems to improve performance. It describes lead, lag, and lag-lead compensators and how they are used to improve transient response, stability, and reduce steady-state error. The benefits of proportional, integral, differential, and PID controllers are outlined. Root locus analysis and design techniques for lead, lag, and PID compensators are presented using examples. Electronic circuit implementations of lead and lag compensators are also covered.

Uploaded by

Ahmed Elbadrawi
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
215 views

Design of Compensators Using Root Locus Diagram

This document discusses various types of compensators that can be added to control systems to improve performance. It describes lead, lag, and lag-lead compensators and how they are used to improve transient response, stability, and reduce steady-state error. The benefits of proportional, integral, differential, and PID controllers are outlined. Root locus analysis and design techniques for lead, lag, and PID compensators are presented using examples. Electronic circuit implementations of lead and lag compensators are also covered.

Uploaded by

Ahmed Elbadrawi
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 20

AUTOMATIC CONTROL

THIRD POWER Prepared by Dr. Helmy El-Zoghby 2013

3/9/2013

Design of compensators using root locus diagram

3/9/2013

A compensator (filter or controller) is a circuit which added to the control system to improve the system performance.

Different types of controllers


1. Two Position Controller 2. Proportional Controller 3. Integral Controller 4. Differential Controller 5. PI Controller 6. PD Controller 7. PID-Controller 8.Lead compensator 9.Lag compensator 10.Lag-lead compensator 11.Intelligent compensators (fuzzy , neural , genetic , .)

3/9/2013

Benefits of compensators
Lead compensator (as PD-controller) is used to: - improve transient response - improve stability Lag compensator ( as PI-controller) is used to: - reduce or eliminate steady-state error Lag- Lead compensator ( as PID-controller) - used to improve all issues.

3/9/2013

Note:
For a closed loop system:

By placing the closed-loop poles at the desired location -if the new closed-loop poles lies on the old root locus , then k is tuned or using P-controller (amplifier). -if the new closed-loop poles not lies on the old root locus , then reshape the root locus by adding poles/zeros (compensator) to system.
3/9/2013 5

Effect of adding poles on control system response

3/9/2013

Effect of adding zeros on control system response

3/9/2013

Lag and lead compensators concept

The controller transfer function

(s) =

(s + z c ) (s + p c )
c

3/9/2013

Lead compensator

3/9/2013

Lag compensator

3/9/2013

10

Electronic circuit of lead or lag compensator

G c (s) =

v v

(s)

(s) i

R C R C
4 3

(s +
1 2

. (s +

R C
1

)
1

= )

(s + z c ) (s + p c )

R C
2

For lead R1C1> R2C2 For lag R1C1< R2C2


3/9/2013 11

Lead compensator design example


Radar tracking system

3/9/2013

12

draw the uncompensated system root locus and find the damping ratio and natural frequency at k=4 design a suitable compensator using three different methods to rise the natural frequency to 4 rad/sec at the same damping ratio draw the compensated root locus draw the designed controller electronic circuit check the steady state error after adding the controller draw the transient response before and after compensation.

3/9/2013

13

solution
For the uncompensated system

K = p2 p2 = 4 then

= 2 , = 60

and = cos( ) = 0 . 5

3/9/2013

14

Angle of deficiency is:

= 180 o + of all zeros of all poles = 180 + 0 (120 + 90) = 180 210 = 30 0

3/9/2013

15

1-Design of compensator using minimum steadystate error method

3/9/2013

16

Selection of the controller pole and zero

3/9/2013

17

The compensated system T.F is :

K c = p1 p2 p c / 4 * zc = 4.68

3/9/2013

18

Steady-state error before and after compensation at unit ramp input Before compensation
k v = lim sG ( s ) H ( s )
s =0

= lim s.

4 s ( s + 2)

s=0

=2

ss

= 0.5 = 50%
v

After compensation
kv
new

= lim sGc ( s )G ( s ) H ( s )

s =0

4.68( s + 2.9) 4 = lim s. ( s + 5.4) s ( s + 2) 1 new ess = new = 0.2 = 20%

s=0

= 5.02

3/9/2013

19

Note : If the input is unit step : Before compensation

k e

= lim =

ss

1+ k

4 = s ( s + 2) s=0 1 = 0%
p

After compensation

k e

new p

= lim =

new

ss

1+k

4.68(s + 2.9) 4 (s + 5.4) s(s + 2) 1 = 0% new


p

s =0

3/9/2013

20

Root locus before and after compensation

3/9/2013

21

Compensator electronic circuit

( s + 2 .9 ) ( s ) = 4 . 68 = ( s + 5 .4 )

R C R C
4 3

(s +
1 2

. (s +

R C
1

)
1

R C
2

)
2

4 . 68 take

R C R C
4 3 3

1 2

, take

= 10 uf

= 10 k

then

= 46 . 8 k ,

= 34 . 5 k ,

= 18 . 5 k
22

3/9/2013

2-Design of compensator using zero-pole cancellation method


Since the Angle of deficiency =300 and

(s) =

(s + z c ) (s + p c )

-Select the controller zero equal to the nearest pole of the system to the desired point to cancel this pole -Find the controller pole from angle condition -Find kc from magnitude condition

3/9/2013

23

Take Zc=2 and from angle condition Pc=4 From magnitude condition

jw p
30

K c = p1 p c / 4 = 4 * 4/4 = 4
The controller T.F become :
G
c

(s) =

(s + 2 ) (s + z c ) = 4 (s + p c ) (s + 4 )

x
-4 -2

x o

-1

x
0

Steady-state error before design


k v = lim s. 4 s ( s + 2)
s =0

= 2, ess =

= 0.5 = 50 %
v

Scale 1:2

Steady-state error after design

kv

new

= lim s.

4( s + 2 ) 4 ( s + 4) s ( s + 2)

s =0

= 4, e ss

new

new v

= 0.25 = 25 %
24

3/9/2013

Design of compensator using PD-controller

PD-controller block diagram

PD-controller electronic circuit

k k

= RC =

R R

2 1

PD-controller transfer function


T .F =
3/9/2013

s =

(s +

k k

p d

) =

(s +

)
25

Since the Angle of deficiency =300


T .F =

s =

(s +

k k

p d

) =

(s +

-the angle of controller zero must equal to 300 -draw a line of 300 with real axis from point (p) -Find the controller zero from angle condition -Find kd from magnitude condition

3/9/2013

26

jw p

Zc=7.8

30

x
-2

-1

x
0

Scale 1:2

Compensated and uncompensated root locus

3/9/2013

27

From angle condition


=

k k

p d

= 7 .8

From magnitude condition

PD-controller electronic circuit

= p1

p 2 / 4 z c = 4 * 3.5/4 * 6.3 = 0.6 = R * C then R = 60 k


2

take C = 10uf then takeR

k
1

R R = 1 k then R
p

= 4 .6 =

= 4 .6 k

3/9/2013

28

The controller T.F become :


T .F =

s = 4 .6 + 0 .6 s

Steady-state error before design


k v = lim s. 4 s ( s + 2)
s =0

= 2, ess =

= 0.5 = 50 %
v

Steady-state error after design

kv

new

= lim s.(4.6 + 0.6s).

4 s(s + 2)

s =0

= 9.2, ess

new

new v

= 0.109 = 10.9%

3/9/2013

29

Using zero-pole cancellation method try to solve this problem and simulate by matlab
A-If it required to design a lead controller to get overshoot=2% and peak time equal 1 sec

Check these values by calculations Lead angle =96 degree , wd=3.14 , wn=5, zeta=.779 ,beta= 38.
1.4

1.2

res po nc e

0.8

0.6

0.4

0.2

3/9/2013

5 Time

10

30

B-at 1 sec peak time and 10% over shoot Check these values by calculations zeta=.57 , wd=3.14 , wn=3.89 , lead angle =35, beta=55
1.4

1.2

res ponse

0.8

0.6

0.4

0.2

5 Time

10

3/9/2013

31

Proportional-Integral controller (lag compensator)


Time output equation is :

p(t ) = k p e(t ) + k i e(t )dt

where Kp is the proportional constant and Ki is the integral constant.

The block diagram of the PI controller

The PI - controller electronic circuit.


32

Design of lag compensator (PI-controller)


The PI-controller transfer function is :

T .F =

k
s

s + s

s + =

k (
p

k k
s

i p

)= k

s +
c

s +

z p
i p

c c

x
zc

pc

The pole of the controller at s=0 The zero of the controller at

s =

k k

3/9/2013

33

PI-controller design steps


The pole of the controller at s=0 The zero of the controller at s=-ki/kp is obtained from angle condition The controller gain constant kp is obtained from magnitude condition

Example
For the following open loop transfer function design a controller to satisfy 0.8 damping ratio , natural frequency of 5 rad/sec , and elimination of steady-state error

G (s)H (s) =
3/9/2013

1 ( s + 3 )( s + 7 )
34

= 5, = cos 1 ( ) = cos 1 ( 0 .8) = 37 0

The steady-state error is zero , PI controller is used


=

T .F

k
(

k
s

s + s

k
s +

s + =

k k
s

i p

)= k

s +

z p

c c

Angle of deficiency is:


= 180
o

of

all zeros

of

all poles

= 180 + 0 (108 + 45 ) = 27
Lag compensator is required
3/9/2013 35

From angle condition

Z
From magnitude condition

k k

i p

= 2 .5

Kc = then

= p 1 p 2 p c / 1 * z c = 20

= 2 . 5 * k p = 50

The controller transfer function is

s +

G
3/9/2013

(s) =

k (
p

k k
s

i p

)=

20

s + 2 .5 s
36

PI-controller electronic circuit


k R
p

R R R

2 1

= 20

take
1

= 1k
2

then

= 20 k

1 RC R

= 50

take C = 10 uf then = 2k
37

3/9/2013

Jw p

Wn=5 27 45 X -7 108 37 X O -3 Zc=-2.5 X Pc=0 Re

Scale 1:1

3/9/2013

38

Thank You & Any Questions?

3/9/2013

39

You might also like