Midterm 3 Solved Problems - Feedback System Analysis and Design | ECE 38200, Exams of Information Systems Analysis and Design

Material Type: Exam; Class: Feedback System Analysis And Design; Subject: ECE-Electrical & Computer Engr; University: Purdue University - Main Campus; Term: Fall 2009;

Typology: Exams

Pre 2010

Uploaded on 12/10/2010

beebcon
beebcon 🇺🇸

5

(1)

20 documents

1 / 4

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
11/16/09
ECE 382 Midterm 3 Solution
Problem 1 (30 points) In the above system, the plant transfer function is G(s) = 1
(s+2)2.
(a) (5 pts) For the uncompensated system with C(s) = 1, find the dominant closed-loop pole(s), the settling
time ts(5% criterion), and the steady-state error for tracking unit step input.
Solution: The system has characteristic equation 1 + G(s) = 0, which is equivalent to (s+ 2)2+ 1 = 0. So
the dominant closed-loop poles are 2±j. The settling time ts=3
2= 1.5 s. The system is of type 0, with
Kp=G(0) = 0.25. Thus, ess = 1/(1 + 0.25) = 0.8.
(b) (5 pts) Given the two design objectives: (i) damping ratio ξ=3
2; (ii) settling time (5% criterion) is
ts= 1 second, find the desired dominant closed-loop poles locations sd.
Solution: sd=σd±d=ξωn±jp1ξ2ωn. Since ts=3
ξωn=3
σd= 1, we must have σd= 3. On the
other hand, ωdd=p1ξ2 = 1/3, which implies ωd=σd/3 = 3. Thus, the desired closed-loop
poles are sd=3±j3.
(c) (12 pts) Design a lead compensator of the form
C1(s) = K1
s+ 3
s+p
by finding K1and pso that the compensated system has dominant closed-loop poles at sdcomputed in (b).
Solution: We first compute the angle of defficiency:
φ= 180[26(sd+ 2)] = 420= 60.
We need to choose pso that it and 3 span φ= 60angle with respect to sd=3 + j3 (only one of
the two desired closed-loop poles is used for convenience). Since 3 is directly below sd, this is indeed very
easy:
p=z+3tan(φ) = 3 + 3 = 6.
To find K1, we used the magnitude condition:
1 + K1
sd+ 3
sd+ 6
1
(sd+ 2)2= 0 K1= 8.
To sum up, the lead compensator is C1(s) = 8s+3
s+6 .
(d) (8 pts) Design a compensator C(s) to satisfy not only the two design objectives in part (b), but also:
(iii) the steady-state error for tracking unit step input is ess <0.1.
Solution: We can add a lag compensator to the lead compensator designed in part (c). For the system
compensated by C1(s), the open-loop transfer function is
C1(s)G(s) = 8(s+ 3)
(s+ 6)(s+ 2)2.
1
pf3
pf4

Partial preview of the text

Download Midterm 3 Solved Problems - Feedback System Analysis and Design | ECE 38200 and more Exams Information Systems Analysis and Design in PDF only on Docsity!

11/16/

ECE 382 Midterm 3 Solution

Problem 1 (30 points) In the above system, the plant transfer function is G(s) = (^) (s+2)^12.

(a) (5 pts) For the uncompensated system with C(s) = 1, find the dominant closed-loop pole(s), the settling time ts (5% criterion), and the steady-state error for tracking unit step input. Solution: The system has characteristic equation 1 + G(s) = 0, which is equivalent to (s + 2)^2 + 1 = 0. So the dominant closed-loop poles are − 2 ± j. The settling time ts = 32 = 1.5 s. The system is of type 0, with Kp = G(0) = 0.25. Thus, ess = 1/(1 + 0.25) = 0.8.

(b) (5 pts) Given the two design objectives: (i) damping ratio ξ =

√ 3 2 ; (ii) settling time (5% criterion) is ts = 1 second, find the desired dominant closed-loop poles locations sd. Solution: sd = −σd ± jωd = −ξωn ± j

1 − ξ^2 ωn. Since ts = (^) ξω^3 n = (^) σ^3 d = 1, we must have σd = 3. On the

other hand, ωd/σd =

1 − ξ^2 /ξ = 1/

3, which implies ωd = σd/

  1. Thus, the desired closed-loop poles are sd = − 3 ± j

(c) (12 pts) Design a lead compensator of the form

C 1 (s) = K 1

s + 3 s + p

by finding K 1 and p so that the compensated system has dominant closed-loop poles at sd computed in (b). Solution: We first compute the angle of defficiency:

φ = 180◦^ − [− 26 (sd + 2)] = 420◦^ = 60◦.

We need to choose −p so that it and −3 span φ = 60◦^ angle with respect to sd = −3 + j

3 (only one of the two desired closed-loop poles is used for convenience). Since −3 is directly below sd, this is indeed very easy:

p = z +

3 ∗ tan(φ) = 3 + 3 = 6.

To find K 1 , we used the magnitude condition:

1 + K 1

sd + 3 sd + 6

(sd + 2)^2

= 0 ⇒ K 1 = 8.

To sum up, the lead compensator is C 1 (s) = 8 s s+3+.

(d) (8 pts) Design a compensator C(s) to satisfy not only the two design objectives in part (b), but also: (iii) the steady-state error for tracking unit step input is ess < 0 .1. Solution: We can add a lag compensator to the lead compensator designed in part (c). For the system compensated by C 1 (s), the open-loop transfer function is

C 1 (s)G(s) =

8(s + 3) (s + 6)(s + 2)^2

It is easy to see that the static error constant is Kp = lims→ 0 C 1 (s)G(s) = 1. Thus, ess = (^) 1+1^1 = 0.5, which

does not satisfy requirement (iii). Indeed, we need to have ess,new = (^) 1+K^1 p,new < 0 .1, i.e., Kp,new > 9. For example, we can set the goal for Kp,new = 10. Then the lag compensator

C 2 (s) = s + 0. 1 s + 0. 01

can be used. To sum up, the compensator that achieves all three design objectives is a lag-lead compensator given by

C(s) = C 1 (s)C 2 (s) = 8

(s + 0.1)(s + 3) (s + 0.01)(s + 6)

Problem 2 (20 points) In the above system, the plant has an unknown stable transfer function G(s). Suppose the plant is first isolated from other parts of the system, and subject to the input signal of 5 sin(20t). By measuring the plant output, it is found that the plant has the steady-state response of 7.5 cos(20t).

(a) (10 pts) Find the plant’s frequency response G(jω) at ω = 20. Solution: The steady-state output is 7.5 cos(20t) = 5 ∗ 1 .5 sin(20t + 90◦). Therefore,

G(j20) = 1. 56 90 ◦^ = j 1. 5.

(b) (10 pts) Suppose the closed-loop system is stable under the compensator C(s) = (^) s+20^10. Determine the steady-state output yss(t) of the closed-loop system under the input u(t) = 2 sin(20t). Solution: The closed-loop system has transfer function

C(s)G(s) 1 + C(s)G(s)

which is assumed to be stable. Thus, the frequency response of the closed-loop system at ω = 20 is

C(j20)G(j20) 1 + C(j20)G(j20)

10 j20+20 ·^ j^1.^5 1 + (^) j20+20^10 · j 1. 5

As a result, under the input u(t) = 2 sin(20t), the steady-state output is

yss(t) = 2 ∗ 0 .3721 sin(20t + 29. 74 ◦) = 0.7442 sin(20t + 29. 74 ◦)

At ω 2 = 0.1, 20 log |G(jω 2 )| ≃ −60 db, and 6 G(jω 2 ) ≃ − 90 ◦.

At ω 3 = 10, 20 log |G(jω 3 )| ≃ −60 db, and 6 G(jω 3 ) ≃ 90 ◦.

At ω 4 = 1000, 20 log |G(jω 3 )| ≃ −60 db, and 6 G(jω 3 ) ≃ − 90 ◦.

Magnitude (dB)

10 −4^10 −3^10 −2^10 −1^100 101 102

0

45

90

Phase (deg)

Bode Diagram

Frequency (rad/sec)