Scribd
Upload
162 views6 pages

Solution To Homework Assignment 4

This document provides solutions to homework problems involving control systems analysis and design. Problem 1 involves analyzing the stability of a closed-loop system using the Routh array and selecting a gain value to minimize overshoot. Problem 2 involves analyzing the stability of another closed-loop system for different parameter values. Problem 3 determines the closed-loop transfer function for a state space system, finds stability conditions, selects parameters for a specified settling time, and plots the step response.

Uploaded by

cavanzas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
162 views6 pages

Solution To Homework Assignment 4

This document provides solutions to homework problems involving control systems analysis and design. Problem 1 involves analyzing the stability of a closed-loop system using the Routh array and selecting a gain value to minimize overshoot. Problem 2 involves analyzing the stability of another closed-loop system for different parameter values. Problem 3 determines the closed-loop transfer function for a state space system, finds stability conditions, selects parameters for a specified settling time, and plots the step response.

Uploaded by

cavanzas
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

Solution to Homework Assignment 4


1. Consider the open-loop transfer function
G(s) =

K(s + 1)
.
(s 1)(s2 + 8s + 32)

(1)

(a) Use the Routh array to determine the range of values of K for which the closed-loop
system with negative unity feedback is stable.
Solution The denominator of the closed-loop transfer function is the numerator of
1 + G(s), which numerator is

p(s) = (s 1)(s2 + 8s + 32) + K(s + 1) = s3 + 7s2 + (24 + K)s + (K 32).

(2)

The Routh array is then


s3
s2
s1
s0
where
=

:
:
:
: (K

1
(24 + K)
7
(K 32)

0
32)

1
1
(K 32 7(24 + K)) = (6K + 200)
7
7

(3)

(4)

and the first element in the s0 row can be computed easily by observing that
1
(0 (K 32)) = K 32.

(5)

Applying the Routh-Hurwitz criterion, the closed loop system is stable if K > 32
(from the s0 row) and K > 200/6 (from the s1 row), so we need K > 32.

(b) Now select a value of K that produces the least overshoot and plot the step response
of the closed-loop system. Be sure to use a time range that shows the important
aspects of the behavior.
Solution:
Using the Matlab script below, we find that the value of seems to
decrease with increasing K, which would indicate that to decrease the overshoot we
should pick a small K. Of course, we have to figure out what value of small is
appropriate. In the matlab script, I tested the value K = 1, and since that yields no
overshoot, we cant do better than that, so K = 1 will do. The plots are shown in
Figures 1 and 2. Heres the script.
%%%
%%% Homework 4, Problem 1 Solution
%%%

sk 10/29/08

for index = 1:1000,


rs(index,:) = roots([1 7 (24 + 32+index) index]);

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

z(index) = -real(rs(index,1))/abs(rs(index,1));
end;
figure(1)
plot(z);
title(HW4 P1: Damping Coefficient \zeta vs. Gain K)
xlabel(Gain K)
ylabel(Damping Coefficient \zeta)
grid
print -depsc HW4_p1_fig1
[y,t] = step(tf([1 1],[1 7 25 1]));
figure(2)
plot(t,y);
title(HW4 P1: Step Response for K=1)
xlabel(Time t)
ylabel(Output y(t))
grid
print -depsc HW4_p1_fig2
HW4 P1: Damping Coefficient vs. Gain K
0.5

0.45

0.4

Damping Coefficient

0.35

0.3

0.25

0.2

0.15

0.1

0.05

100

200

300

400

500
Gain K

600

700

800

900

1000

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

HW4 P1: Step Response for K=1


1

0.9

0.8

0.7

Output y(t)

0.6

0.5

0.4

0.3

0.2

0.1

50

100
Time t

150

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

2. Consider a unity negative feedback system with forward path (open loop) transfer function
G(s) =

K
.
s((s + 3 + K/3)(s + p) + (K/3))

(6)

(a) Draw a block diagram, in which each block contains a transfer function having at
most one pole and at most one zero that would implement this closed-loop system.
(There is more than one way to do this. Any correct answer will do.)
(b) Assume that the value p may take any value in the set {1, 2, 3}. Find the range of
values for the gain K that result in a stable closed-loop system regardless of which
of the three values p takes.
Solution: A little tedious algebra yields the Routh array:
s3 :

s2 :
s1 :
s0 :

9+3p+K
3
9+3p2K
9+3p+K

3+

K
3

p+

K
3

K
0

(7)

which gives us three constraints on the value of K for the system to be stable, namely
9 + 3p + K > 0

(8)

9 + 3p 2K > 0

(9)

K > 0.

(10)

The first condition simplifies to K > 9 3p. For the three given values of p, the
maximum of the right hand side occurs when p = 1, yielding K > 12. The second
condition simplifies to K < (9 + 3p)/2. The minimum of the right hand side occurs
when p = 1, yielding the constraint K < 6. Thus, the acceptable range of values for
the gain K is 0 < K < 6.
3. Consider the Single-Input Single-Output (SISO) system with state space representation
x(t)

=
y(t) =
and input u(t) = Kx(t) + r(t).

"
h

0
2
1 3
1 0

x(t) +

"

0
1

u(t)

(11)

x(t)

(12)

(a) Determine the closed loop transfer function y(s)/r(s).


Solution: First, we substitute for u in the state equation to obtain
x(t)

= Ax(t) + B (Kx(t) + r(t))

(13)

= (A BK)x(t) + Br(t)
"

=
=

"

0
1

0
2
1 k1 3 k2

0
2
1 3

"

= Acl x(t) + Br(t).

(14)
k1 k2

x(t) +

"

!
i

0
1

x(t) +
r(t)

"

0
1

r(t)

(15)
(16)
(17)

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

The transfer function is then


g(s) =

y(s)
r(s)

= C (sI Acl )1 B + D
=
=

1 0

"

(18)

s
2
1 + k1 s + 3 + k2

#1 "

0
1

+0

(19)

2
2
= 2
. (20)
s(s + 3 + k2 ) (2(1 k1 ))
s + (3 + k2 )s + 2(k1 1)

(b) Determine the range of values k1 and k2 for the closed-loop transfer function to be
stable.
Solution: Using the Routh Array or any other valid method we find that we need
both 3 + k2 and 2(k1 1) positive, or k2 > 3 and k1 > 1.
(c) Select values of k1 and k2 such that the settling time of the unit step response is 1
second.
Solution:
The problem should have specified which settling time was required,
but we know that if this is not specified, the 2%-settling time is usually a reasonable
choice. Then we want 1 = Ts = 4n , so we need n = 4 and the ki must satisfy
3 + k2 = 2n = 8
2(k1 1) =

n2 .

(21)
(22)

Obviously we need k2 = 5. We have to think a little more to choose an appropriate


value for k1 . It might appear that we could choose any n whatsoever, but for a step
response we would generally want to use a around 0.707. To get a quick order of
magnitude estimate, I observe that for = 1 wed need k1 = 9.
Decreasing would
increase n and thus the required
value for k1 . Since 0.707 = 1/ 2, decreasing to

90.707 would increase nby 2, hence n2 by 4. We thus need to roughly double k1 .


To be precise, for = 1/ 2 we need k1 = 17.
(d) For your matrix K, calculate the step response of the state space system. (That
means you need to determine both x(t) and y(t)).
Solution:
Having been given no initial condition we assume x(0) = 0. We note
that the output is just y = x2 so if we determine the step response of the states we
will also have the step response of the output. I used the Matlab script below to
obtain the step response that follows.
%%%
%%% Homework 4, Problem 3 Solution
%%%
k1 = (4/0.707)^2/2+1
k2 = 5;
A = [0 2;1-k1 -3-k2];
B = [0;1];
C = eye(2);
D = 0*B;

sk 11/15/08

ECE382/ME482 Fall 2008

Homework 4 Solution

November 15, 2008

[y,t] = step(ss(A,B,C,D));
figure(2)
plot(t,y);
title(HW4 P3: Step Response for K=1)
xlabel(Time t)
ylabel(States)
legend(x_1,x_2)
grid
print -depsc HW4_p3_fig1
HW4 P3: Step Response for K=1
0.09
x
x

0.08

1
2

0.07

0.06

States

0.05

0.04

0.03

0.02

0.01

0.01

0.5

1
Time t

1.5

You might also like