Chapter Objective: - Plot and Sketch The Frequency Response of A System - Nyquist Plots - Nyquist Stability Criterion - Relative Stability

Chapter 10
Frequency Response Nyquist Plot
Chapter Objective:
Plot and sketch the frequency response of a system
Nyquist Plots
Nyquist Stability Criterion
Relative Stability
Feb-17 ME2142 Feedback Control Systems C10.1
Copyright ME, NUS. All Rights Reserved.
Frequency Response of a System
From previous chapter, we have:
Given a system ,
U(s) Y(s)
G(s)
the frequency response of the system is

(1)
We are interested the gain
(2)
and phase shift,
(3)

Polar Plot (Nyquist Plot)
Given the frequency response, , we
can express it in terms of its magnitude
Im
and its angle .
The plot of versus on
the polar co-ordinate with varying
frequency is called the Nyquist Plot. Re

For a given frequency, :
Magnitude: a distance from the
origin
Angle: an angle with respect
to the positibe real axis. (i.e. A positive
(negative) phase angle is measured
counter-clockwise (clockwise) from the
positive real axis)
As we have lossed the frquency
information, we usually label the point with
its frquency.

Simple Polar Plots
Integrator and Derivative terms
Im
Integrator, , we have 1
Re
90
and 90
or Arrowindicate
directionoffrequency
0 increasing.
0,
Differentiator, , we have Im
and 90
or
0 90
0, Re

Simple Polar Plots - First Order Lag
First order Lag term.
, The frequency response is
Im
0.5 1.0
and tan Re
or
At 0, 0.5
1
0 1 and 0 0
T
At ,
and tan 1 45
As ,
0 and tan 90
The polar plot of this transfer function is a semicircle.

Is 1st Order Lag FR a Semicircle?
We have
(*)
and

1
0
0.25 0.25
0.5 0.5 Equationofacirclewithcenterat
(0.5,0)

Simple Polar Plots - First Order Lead
First order Lead term.
1 , The frequency response is
1
1 and tan
Im
i.e.
1 and 1
1.0
At 0, T
0 1 and 0 0
At ,
Re
1 1 2 and tan 1 45 1.0
As ,
and tan 90

Polar Plots 2nd Order Lag
Consider a system
Im
,
The frequency response is

1
0
Re
where
:Large

Therefore
and tan
At 0 (or 0),
0 1 and 0 0
As (or ),
0 and tan 180
:Small
At (or 1),
and tan 90

Polar Plots 2nd Order Lag
Resonant frequency of 2nd order systems
Polar plots at different values of

2nd Order numerator term
Consider a system

Im
2 1,
The frequency response is
2 1 1 2
where
Re

0 1
Therefore
1 4 and tan
At 0 (or 0),
0 1 and 0 0
As (or ),
and 180

Example 10.1
Consider the following second-order transfer function:
1
1
Sketch a polar plot of this transfer function.
Frequency response can be written as
We have
1
1
and
90 tan

Example 10.1
We have
1 1
Im
1 1 1
and its gain and phase are
1
-T
1
Re
and
90 tan
At 0,
lim

i.e. and 90
At ,
0 and 180

General Shape of Polar Plots
The polar plots of a transfer function of the
form

For 0 or Type 0 system:
At 0, point is finite and is on the real
axis
At , the FR locus tends towards the
origin at an angle of 90
At 0, , 90
At , 0, 90
At 0, , 180
At , 0, 90

Effects of numerators
Numerator dynamics can complicates polar plot curves
Example:
Consider a system
1
1 1
Imaginary Axis
Imaginary Axis
Imaginary Axis
Real Axis
Real Axis Real Axis
1 1 1 1 1 1 1 1 1

Samples of Polar plot
Extracted from Ogata Modern Control Engineering

Nyquist Stability Criterion
Consider the closed-loop system shown on the right.
The closed-loop transfer function is Rs Ys
G(s)
. +
-
The closed-loop characteristic equation is
H(s)
1 0.
We know that, for stability of closed-loop system, all
roots of the closed-loop characteristic equation must lie
in the left-half plane.
Nyquist Stability Criterion relates the open-loop
frequency response
to the number of zeros and poles of

1
that lie in the right-half plane.

Some mathematical background
How do we visualize a complex function?
Example:

For ,

2
2 cos 2 2cos sin 2 2sin
where

Aug-15 ME2142 Feedback Control Systems C2.17

Some mathematical background
Consider a complex function 22
s sPlane
Plane
1 11
22
where is a complex variable.
Imaginary Axis
Imaginary Axis
11
00
R
Let us define a closed path, , on
the s-plane such that 11
,0 2 . 22
3 3 2 2 1 1 00 1 1 22 33
Hence, we have Real

RealAxis
Axis
1 33
FF Plane
Plane
is known as a function that maps 22
FF2
the complex variable from the s- 11
2
Axis
plane to the F-plane.
ImaginaryAxis
FF1
1
00
Imaginary
Case 1: 0.5 1
11
No encirclement of origin
22
Case 2: 2 1
CCW encirclement of origin 33
3
3
2
2
1
1
0
0
1
1
2
2
3
3
Real
RealAxis
Axis

Mathematical background
Consider another complex function 2
s Plane 1
1
2
Imaginary Axis
With the same closed path, , on 0
the s-plane
1
We have
2
3 2 1 0 1 2 3
Real Axis
Case 1: 0.5 1 3
F1 move in CW direction F Plane
2
No encirclement of origin
1 F2
Imaginary Axis
F1
0
Case 2: 2 1
CW encirclement of origin 1
3
3 2 1 0 1 2 3
Real Axis

Consider another complex function
s Plane
Let be a closed path on the s-plane
Imaginary Axis
defined by
: ,0 2
Case I: 0.5
In s-plane: Real Axis
did not encircle any pole or zero.
In F-plane: 1
We have 0.8
0.6
2 0.4
3 1
Imaginary
0.2
No encirclement of 0
0.2
origin. 0.4
CCW 0 0.3 0.6

0.8
/2 0.4 0.5 1
2.5 2 1.5 1 0.5 0
2 Real
3 /2 0.4 0.5
s Plane
With defined by
Imaginary Axis
: ,0 2
Case II: 1.5
In s-plane:
did not encircle any pole or zero. 5 4 3 2 1 0 1
Real Axis
2 3 4 5
In F-plane: 1.5
Encirclement of origin. 1
CW
0.5
Imaginary 0
0.5
0 0
/2 0.2 0.4 1
4.7 1.5
1 0 1 2 3 4 5
3 /2 0.2 0.4 Real

s Plane
With defined by
Imaginary Axis
: ,0 2
Case III: 2.5
In s-plane:
did not encircle any pole or zero. 5 4 3 2 1 0
Real Axis
1 2 3 4 5
In F-plane: 2.5
No encirclement of origin. 2
1.5
1
0.5
Imaginary 0
0 0.026 0.5
/2 0.3 0.1 1
1.5
6
2
3 /2 0.3 0.1 2.5
0 1 2 3 4 5 6 7
Real

s Plane
With defined by
Imaginary Axis
: ,0 2
Case III: 4
In s-plane:
did not encircle any pole or zero. 5 4 3 2 1 0
Real Axis
1 2 3 4 5
In F-plane: 1.5
CW encirclement of origin. 1
0.5
Imaginary 0
0 0.1
0.5
/2 0.21 0.05
2 1
3 /2 0.21 0.05 1.5

2.5 2 1.5 1 0.5 0 0.5
Real

Mapping Theorem
Given that is a rational function that maps the complex variable
from the s-plane to the F-plane. and are the polynomials of the
numerator term and denominator term, respectively. And , where is a
closed path traveling in the clockwise direction.
Then, the locus will encircle the origin in an anticlockwise direction
times, where
= number of poles encircled by , and
= number of zeros encircled by .
Example: Consider a function,

.
.
. . .
Say, we define as ,0 2 .
Lets try to verify this trying out different value of R.

Case 1:
1.5
1
splane
Imaginary
3 0.5
0
Fplane 0.5
2 1
1.5
2 1 0 1 2 3
1 Real
Imaginary
2 CCW encirclements on origin

3
3 2 1 0 1 2 3
Real

Case 2:
0.4
splane
0.2
Imaginary
25
0
20
Fplane
0.2
15
0.4
0.4 0.2 0 0.2 0.4 0.6
10
Real
5
Imaginary
10
15
20
25
20 10 0 10 20 30 40 50
Real

Case 3:
0.2
splane
Imaginary
80
0
Fplane
60
0.2
40 0.2 0 0.2 0.4
Real
20
Imaginary
20
40
60

80
120 100 80 60 40 20 0 20 40 60
Real

Case 4:
0.2
splane
0.1
Imaginary
50
0
40
Fplane
0.1
30
0.2
0.2 0.1 0 0.1 0.2 0.3
20
Real
10
Imaginary
10
20
30
40
No CCW encirclements on origin
50
0 20 40 60 80 100 120 140 160
Real

Can mapping theorem be used for stability analysis?
Application of the mapping theorem to stability analysis of closed-loop system.

Question:
Can we make use of this mapping theorem to analyze the closed-loop
stability of the system from open-loop information?
Answer:
Yes, we can, if we have solved these two problems:
i. Define an appropriate .
ii. Define an appropriate .

(i) Definition of F(s)
Rs Ys
Consider a generic closed-loop system as shown G(s)
on the right. +
-
Open-loop transfer function considered here is .
H(s)
We can deduce that the closed loop transfer function is
We can use the closed loop characteristic equation as as such

1
Thezerosof arealsotherootsoftheclosed
1 loopcharacteristicequation(orclosedlooppoles)
ThepolesofF(s)arealsotheopenloop
polesofthesystem.
If we can find a closed-contour such that it encircle all the unstable roots of
the closed loop characteristic equation, then we can tells the stability of the
systems by just observing the number of encirclement on the origin.

(ii) Finding an appropriate .
Im
Definition of : 2
We need a such that it covers the entire

right-half plane of . One solution is to define
as follows: R
1 2 , the contour moves from the origin along the
positive imaginary axis until it reaches the infinity. 1 Re
2 3 , it then moves on a large semi-circle of radius 4
. must be chosen large enough that it covers all the
unstable poles and zeros of . This will swing the
contour from to .
3 4 , this part of the contour will moves from
back to the origin.
3
Hence, with
,0 Say,theDcontourdefinedabovehasencircled
numberofzeros,and
, , with numberofpoles
0 ofthefunction .Then,thenumberof
This contour is called the D-contour. clockwiseencirclementsoftheorigininFplane
willbe

Summary: Pre-Nyquist Stability Criterion
Im
For a closed-loop system as shown below. S-plane
Rs Ys
G(s) Re
+
-
H(s)
The stability of the closed-loop system can be
Im
determined by observing on the number
encirclement of the origin for F-plane
1
for on the D-contour.
The number of unstable closed-loop poles ,
can be determined by Re
where is the number of clockwise

encirclement of the origin in GH-plane and is
the number of unstable open-loop poles of
.
Examples
Case I: Im
Observation: F-plane
System is open-loop stable ( 0), and
shows two time of clockwise encirclement of
the origin ( 2). Re
Conclusion:
2 0 2
System closed-loop unstable (2 unstable
closed-loop poles)
Case II:
Im
System is open-loop stable ( 0), and
shows zero of clockwise encirclement of the
origin ( 0).
Re
Conclusion:
0 0 0
System closed-loop stable

Examples
Case III: Im
F-plane
Observation:
System is open-loop unstable with one
unstable open-loop pole ( 1), and
Re
shows zero clockwise encirclement of the
origin ( 0).
Conclusion:
0 1 1
System closed-loop unstable (1 unstable
closed-loop poles)
Case IV: Im
System is open-loop unstable with one

unstable open-loop pole ( 1), and
shows one anti-clockwise encirclement of the Re
origin.( 1).
Conclusion:
1 1 0
System closed-loop stable
Graphical interpretation of Nyquist Stability Criterion
Given an open-loop transfer function, and its closed characteristic

equation F 1 . With , we have
Im Im
F-plane GH-plane
Re -1 Re
1
Therefore, instead of plotting 1 , we use plot of .

If we are plotting , then we are looking for encirclement of the 1
point rather than the origin.
Interestingly, this is also the frequency response of the open-
loop system
Stability Criterion Im
S-plane
For a closed-loop system as shown below.
Rs Ys Re
G(s)
+
-
H(s)
The stability of the closed-loop system can be determined
by observing on the number encirclement of the -1 point GH-plane
Im
of for on the D-contour.

The Nyquist Stability criterion is expressed as
-1 Re
where
Z = number of unstable closed-loop poles
N = the number of clockwise encirclement on the
1 point
P = the number of poles of G(s)H(s) in the right-half
plane.

Example 1
Given
1 1
Gain is Im
1 1
Phase is
tan tan
When 0, 0
When , 0 180
Re
Nyquist condition is
We have,
0, 0 0
system is closed-loop stable.

Example 2
Given
2
3 6.25 1
Gain is
2
6.25 3 1
Phase is
3 Im
tan tan tan
2 1 6.25
At 0,
.
At , 0 180
At 1, 1 1 0.26 138 Re
At 5, 5 5 0.439 108
We have,
1, 0 1
system is closed-loop unstable for 1.
How can we make it stable?
Relative Stability
Beside stability, we also like to have some
measure on the degree of the stability Im
of a system.
In here, we introduce two measures of
1
relative stability
Kg
Gain Margin
Gain margin is defined as the reciprocal
Re
of the magnitude at the
frequency where phase angle is
180.
Phase Margin.
180 plus the phase angle of the
systems open-loop transfer function at
the frequency where
1.

Graphical interpretation of GM & PM

Relation between Closed-Loop Transient
and Open-Loop Frequency Responses
Consider a unity feedback system whose open-loop transfer function,
(*)
The closed-loop transfer function of the system will be
(**)
Comparing (**) with a generic 2nd order system of the form , we can
deduce that the O.L.T.F. (*) can be written in the form of
where and are the damping ratio and natural undammed frequency of the
closed-loop system, respectively.

Relation between Closed-Loop Transient
and Open-Loop Frequency Responses
In order to evaluate the phase margin of the open-loop
system, we first find the frequency for which 1.
Hence,
90
1
80
Solving this equation will yields
70
2 1 4
60
PhaseMargin,
Therefore, at this frequency, the phase is 50
40
90 tan
30
Hence the phase margin, , is
20
180
10
180 90 tan 0
0 0.5 1 1.5 2
tan DampingRatio,

Worked Example 1:
Experimental Determination of FR
A system with experimentally determined
frequency response is shown in the table below. (rad/s)
Gain Phase
(dB) ( )
It is controlled by a proportional controller as 0.01 55.01 0.6
shown below. 0.02 48.99 1.1
0.04 42.97 2.3
System 0.08 36.94 4.6
+ 0.1 35.01 5.7
- 0.2 28.98 11.3
0.4 22.96 21.8
Sensor 0.8
1
16.92 38.7
14.96 45.0
2 8.82 63.4
The sensor has a dynamics of 4 2.32 76.0
1 8 -5.20 82.9
10 -8.00 84.3
1 20 -18.00 87.1
40 -29.34 88.6
i) Plot the open-loop frequency response of system.

ii) Find the Gain and Phase Margin of the system.
iii) Hence, find the range of K such that the system is
closed-loop stable.
Worked Example 1: Solution
Solution:
From the dynamics of the
sensor, we have
Hence,
10 log 1 G(s) H(s) G(s)H(s)
Gain Phase Gain Phase Gain Phase
tan 0.01 55.01 90.1 0.00 0.6 55.01 90.6
For the overall gain and phase, 0.02 48.99 90.1 0.00 1.1 48.98 91.3
we have 0.04 42.97 90.2 0.01 2.3 42.96 92.5

0.08 36.94 90.5 0.03 4.6 36.92 95.0
0.1 35.01 90.6 0.04 5.7 34.96 96.3
0.2 28.98 91.1 0.17 11.3 28.81 102.5
0.4 22.96 92.3 0.64 21.8 22.31 114.1
0.8 16.92 94.6 2.15 38.7 14.77 133.2
1 14.96 95.7 3.01 45.0 11.95 140.7
2 8.82 101 6.99 63.4 1.83 164.7
4 2.32 112 12.30 76.0 9.98 187.8
8 5.20 129 18.13 82.9 23.33 211.5
10 8.00 135 20.04 84.3 28.05 219.3
20 18.00 153 26.03 87.1 44.04 240.6
40 29.34 166 32.04 88.6 61.38 254.5

Plotting of 6060
60 180
180
180
Plotting of 4040
40 135
135
135
Finding Gain Margin
2020
20 GM
GM 9090
90
6
6 6dB
00 0
6dB
4545
45
Finding Phase Margin 2020

20 00 0
Phase(Deg)
Phase(Deg)
Phase(Deg)
Gain(dB)
Gain(dB)
Gain(dB)
168
4040
40 4545
45
12
6060
60 90
90
90
8080
80 135
135
135
168
100
100
100 180
180
180
120
120
120 225
225
225
140
140
140 270
270
270
0.01
0.01
0.01 0.1
0.1
0.1 11 1 1010
10 100
100
100
Frequency(Rad/s)
Frequency(Rad/s)
Frequency(Rad/s)

Worked Example 2:
Example 2:
The open-loop transfer function of a unity feedback control system is
1 0.1 1 0.001
Determine the value of if the steady-state error of the output position must be
less than or equal to 0.1 to a unit ramp input. With this value of , what is the
gain margin and phase margin of the system? Find the resonant frequency
and resonant gain of the system.

Example 2: Solution
The closed-loop transfer function of the closed-loop system is
1
Tracking error, , is
1
1 1
For unit ramp input, ,
lim

lim

0.1
10

For K = 10, the frequency response is 40 180
. . 20 135
20 log
0 90
20 20 log
10 log 1 0.01 20 GM 45
10 log 1 0.000001
90 tan 0.1 tan 0.001 40 0
Phase(Deg)
Gain(dB)
60 45
From the plotted Bode Diagram, the
Gain and Phase margins are 80 90
Gain Margin = 40 dB 100

128
135
Phase Margin = 52 PM
120 180
140 225
160 270
0.1 1 10 100 1000
Frequency(Rad/s)

The End


Chapter Objective: - Plot and Sketch The Frequency Response of A System - Nyquist Plots - Nyquist Stability Criterion - Relative Stability

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter Objective: - Plot and Sketch The Frequency Response of A System - Nyquist Plots - Nyquist Stability Criterion - Relative Stability

Uploaded by

Copyright:

Available Formats

Chapter 10

Frequency Response Nyquist Plot

the frequency response of the system is

Feb-17 ME2142 Feedback Control Systems C10.2

frequency is called the Nyquist Plot. Re

Feb-17 ME2142 Feedback Control Systems C10.3

Feb-17 ME2142 Feedback Control Systems C10.4

Feb-17 ME2142 Feedback Control Systems C10.5

Feb-17 ME2142 Feedback Control Systems C10.6

Feb-17 ME2142 Feedback Control Systems C10.7

The frequency response is

Feb-17 ME2142 Feedback Control Systems C10.8

Resonant frequency of 2nd order systems

Polar plots at different values of

Feb-17 ME2142 Feedback Control Systems C10.9

Feb-17 ME2142 Feedback Control Systems C10.10

Feb-17 ME2142 Feedback Control Systems C10.11

Feb-17 ME2142 Feedback Control Systems C10.12

Feb-17 ME2142 Feedback Control Systems C10.13

Feb-17 ME2142 Feedback Control Systems C10.14

Extracted from Ogata Modern Control Engineering

to the number of zeros and poles of

Feb-17 ME2142 Feedback Control Systems C10.16

Aug-15 ME2142 Feedback Control Systems C2.17

Hence, we have Real

Feb-17 ME2142 Feedback Control Systems C10.18

Feb-17 ME2142 Feedback Control Systems C10.19

Let be a closed path on the s-plane

CCW 0 0.3 0.6

3 /2 0.2 0.4 Real

Feb-17 ME2142 Feedback Control Systems C10.21

Feb-17 ME2142 Feedback Control Systems C10.22

3 /2 0.21 0.05 1.5

Feb-17 ME2142 Feedback Control Systems C10.23

Example: Consider a function,

Feb-17 ME2142 Feedback Control Systems C10.24

2 CCW encirclements on origin

Feb-17 ME2142 Feedback Control Systems C10.25

Feb-17 ME2142 Feedback Control Systems C10.26

2 CCW encirclements on origin

Feb-17 ME2142 Feedback Control Systems C10.27

Feb-17 ME2142 Feedback Control Systems C10.28

Application of the mapping theorem to stability analysis of closed-loop system.

Feb-17 ME2142 Feedback Control Systems C10.29

We can use the closed loop characteristic equation as as such

Feb-17 ME2142 Feedback Control Systems C10.30

We need a such that it covers the entire

Feb-17 ME2142 Feedback Control Systems C10.31

where is the number of clockwise

Feb-17 ME2142 Feedback Control Systems C10.33

System is open-loop unstable with one

Given an open-loop transfer function, and its closed characteristic

Therefore, instead of plotting 1 , we use plot of .

of for on the D-contour.

Feb-17 ME2142 Feedback Control Systems C10.36

Feb-17 ME2142 Feedback Control Systems C10.37

Feb-17 ME2142 Feedback Control Systems C10.39

Feb-17 ME2142 Feedback Control Systems C10.40

Feb-17 ME2142 Feedback Control Systems C10.41

Feb-17 ME2142 Feedback Control Systems C10.42

i) Plot the open-loop frequency response of system.

we have 0.04 42.97 90.2 0.01 2.3 42.96 92.5

Feb-17 ME2142 Feedback Control Systems C10.44