You are on page 1of 16

Jayasinghe H. B. C.

EN17409056
Control Systems Lab 4
Example 1

Consider a unity-feedback system with the following open-loop transfer function, and draw a
Nyquist plot with MATLAB and examine the stability of the closed-loop system.

20(𝑠 2 + 𝑠 + 0.5) 20𝑠 2 + 20𝑠 + 10


𝐺(𝑠) = =
𝑠(𝑠 + 1)(𝑠 + 10) 𝑠3 + 11𝑠 2 + 11𝑠 + 10𝑠

>> num=[0 20 20 10];

>> den=[1 11 10 0];

>> sys=tf(num, den);

>> nyquist(sys);

>> v=[-2 -3 3 3];

>> axis(v);

>> grid
Post Lab Work
Exercise 1

Obtain a Bode plot for the following system and use it to find the gain margin and phase margin.

1 1
𝐺(𝑠) = =
(𝑠 + 1)2 (𝑠 2 + 𝑠 + 4) 𝑠4 + 3𝑠 3 +7𝑠 2 + 9𝑠 + 4

>> num=[0 0 0 0 1];

>> den=[1 3 7 9 4];

>> sys=tf(num,den);

>> margin(sys)

gain margin = 18.1 dB at 1.73 rad/s

phase margin = infinite


Exercise 2

In this exercise you will study the Bode plots of normalized second-order systems with an
additional pole described by,

1
𝐺(𝑠) = 𝑠
(𝑝 + 1) (𝑠 2 + 𝑠 + 1)

a. Draw Bode plots of G(s) with p = α/2 for α = 0.1, 1.

When p = α/2

1
𝐺(𝑠) =
2𝑠
( α + 1) (𝑠 2 + 𝑠 + 1)

α = 0.1,
1 1 1
𝐺(𝑠) = = =
2𝑠 (20𝑠 + 1)(𝑠 + 𝑠 + 1) 20𝑠 + 21𝑠 2 + 21𝑠 + 1
2 3
(0.1 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(1,[20 21 21 1]);

>> margin(G)
α = 1,
1 1
𝐺(𝑠) = =
2𝑠 2𝑠 3 + 3𝑠 2 + 3𝑠 + 1
( 1 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(1,[2 3 3 1]);


>> margin(G)

Do the plots lead you to expect additional rise times?


b. Draw the Bode plots for several (at least seven) other values of α.

α=2

1 1 1
𝐺(𝑠) = = =
2𝑠 (𝑠 + 1)(𝑠 2 + 𝑠 + 1) 𝑠 3 + 2𝑠 2 + 2𝑠 + 1
( 2 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(1,[1 2 2 1]);

>> margin(G)

α=4

1 2 2
𝐺(𝑠) = = =
2𝑠 (𝑠 + 2)(𝑠 2 + 𝑠 + 1) 𝑠 3 + 3𝑠 2 + 3𝑠 + 2
( 4 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(2,[1 3 3 1]);

>> margin(G)
α=6

1 3 3
𝐺(𝑠) = = = 3
2𝑠 2 2
(𝑠 + 3)(𝑠 + 𝑠 + 1) 𝑠 + 4𝑠 + 4𝑠 + 3
( 6 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(3,[1 4 4 3]);

>> margin(G)
α = 10

1 1 1
𝐺(𝑠) = = =
2𝑠 (2𝑠 + 10)(𝑠 2 + 𝑠 + 1) 2𝑠 3 + 12𝑠 2 + 12𝑠 + 10
(10 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(10,[2 12 12 10]);

>> margin(G)

α = 20

1 10 10
𝐺(𝑠) = = = 3
2𝑠 2 2
(𝑠 + 10)(𝑠 + 𝑠 + 1) 1𝑠 + 11𝑠 + 11𝑠 + 10
( + 1) (𝑠 2 + 𝑠 + 1)
20

>> G = tf(10,[1 11 11 10]);

>> margin(G)
α = 100

1 50 50
𝐺(𝑠) = = = 3
2𝑠 2 2
(𝑠 + 50)(𝑠 + 𝑠 + 1) 1𝑠 + 51𝑠 + 51𝑠 + 50
(100 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(50,[1 51 51 50]);

>> margin(G)
α = 1000

1 500 500
𝐺(𝑠) = = =
2𝑠 (𝑠 + 500)(𝑠 2 + 𝑠 + 1) 1𝑠 3 + 501𝑠 2 + 501𝑠 + 500
(1000 + 1) (𝑠 2 + 𝑠 + 1)

>> G = tf(500,[1 501 501 500]);

>> margin(G)

When α increases, bandwidth was increased.


Exercise 3

Obtain a Nyquist plot for the open-loop transfer function,

𝐾
𝐺(𝑠) =
(𝑠 + 1)(𝑠 + 2)(𝑠 + 3)

with K = 20 in the unity-feedback system and determine whether the closed-loop system is stable
using the Nyquist stability criterion.

𝐾 20 20
𝐺(𝑠) = = 2 = 3 2
(𝑠 + 1)(𝑠 + 2)(𝑠 + 3) (𝑠 + 3𝑠 + 2)(𝑠 + 3) 𝑠 + 6𝑠 + 11𝑠 + 6

>> G = tf(20,[1 6 11 6]);

>> nyquist(G)

>> controlSystemDesigner('bode',G);

The Nyquist plot does not encircle the -1 + j0 point, so that, N = 0. And no open-loop poles lie in
the right-half plane, so that., P = 0, implying that Z = N + P = 0. Therefore, closed-loop system is
stable (the right hand side Bode Diagram can also use to check the stability).
Repeat for K = 50,

𝐾 50
𝐺(𝑠) = = 3
(𝑠 + 1)(𝑠 + 2)(𝑠 + 3) 𝑠 + 6𝑠 2 + 11𝑠 + 6

>> G = tf(50,[1 6 11 6]);

>> nyquist(G)

>> controlSystemDesigner('bode',G);

The Nyquist plot does not encircle the -1 + j0 point, so that, N = 0. And no open-loop poles lie in
the right-half plane, so that., P = 0, implying that Z = N + P = 0. Therefore, closed-loop system is
stable (the right hand side Bode Diagram can also use to check the stability).
Exercise 4

For a unity feedback system with the open-loop transfer function,

(𝑠 + 1)(𝑠 + 2)
𝐺(𝑠) =
𝑠 2 (𝑠 + 3)(𝑠 2 + 2𝑠 + 25)

a. Use Bode and root-locus plots to determine the gain and frequency at which a close-loop
pole falls on the imaginary axis.

(𝑠 + 1)(𝑠 + 2) 𝑠 2 + 3𝑠 + 2
𝐺(𝑠) = =
𝑠 2 (𝑠 + 3)(𝑠 2 + 2𝑠 + 25) 𝑠 5 + 5𝑠 4 + 31𝑠 3 + 75𝑠 2

>> G = tf([0 0 0 1 3 2],[1 5 31 75 0 0]);


>> rlocus(G);
>> margin(G)

According to the Bode plot when closed-loop pole falls on the imaginary axis,

Gain ≈ 52.5
Frequency = 4.96 rad/s
According to the root-locus plot when closed-loop pole falls on the imaginary axis,

Gain = 34.4 dB
Frequency = 4.96 rad/s

What gain (or gains) gives a PM of 20°

PM of 20° was obtained when Phase = (20° − 180°) = − 160°


Thus,
Gain = −11.4 dB @ Frequency = 0.324 rad/s ≈ 0.325 rad/s

Gain = −33.8 dB @ Frequency = 4.58 rad/s ≈ 4.6 rad/s

b. What is the gain margin when PM= 20°?

Gain Margin = −11.4 dB − ( − 33.8 dB ) = 22.4 dB


Exercise 5

Suppose you are given the plant transfer function

(𝑠 + 𝛽)
𝐺(𝑠) =
𝑠(𝑠 + 1)(𝑠 2 + 8𝛼 + 52)

In the unity feedback control system, study the behavior of the locus of closed loop poles as the
parameter β is varied from 0 to 10, paying particular attention to the region between 2.5 and 3.5.

Assuming 𝛼 = s

(𝑠 + 𝛽)
𝐺(𝑠) =
𝑠(𝑠 + 1)(𝑠 2 + 8𝑠 + 52)

β=0

(𝑠 + 0) 𝑠
𝐺(𝑠) = =
𝑠(𝑠 3 + 9𝑠 2 + 60𝑠 + 52) 𝑠 4 + 9𝑠 3 + 60𝑠 2 + 52𝑠

>> G = tf([0 0 0 1 0],[1 9 60 52 0]);

>> rlocus(G);
β=5

𝑠+5
𝐺(𝑠) =
𝑠 4 + 9𝑠 3 + 60𝑠 2 + 52𝑠

>> G = tf([0 0 0 1 5],[1 9 60 52 0]);

>> rlocus(G);

β = 10

𝑠 + 10
𝐺(𝑠) =
𝑠 4 + 9𝑠 3 + 60𝑠 2 + 52𝑠
>> G = tf([0 0 0 1 10],[1 9 60 52 0]);

>> rlocus(G);

>> rlocus(G);
Gain & Frequency are reduced when β is increased.

Determine the value of α that a multiple closed-loop pole occurs.

You might also like