NPTEL

Control Engineering
Solution: Assignment 7

Common data for Q1-Q2

A first-order system has the transfer function

1
G(s) =
s+2
With a PI controller the closed loop system (unity feedback) has the characteristic polynomial

p(s) = s2 + 3s + 4.

1. The value of KP = . [Ans=1; Range = 0.95 to 1.05]

2. The value of KI = . [Ans=4; Range = 3.95 to 4.05]
Solution:
Gc (s) = KP + KI /s
C(s) Gc (s)G(s) sKP + KI sKP + KI
= = = 2
R(s) 1 + Gc (s)G(s) s(s + 2) + sKP + KI s + (KP + 2)s + KI

Comparing denominator with s2 + 3s + 4, we get KP = 1 and KI = 4.

Common data for Q3-Q5

Consider a system with the transfer function

G(s) = 1/(s + 1)2 .

3. An integral controller Gc (s) = KI /s gives a closed-loop pole at s = −a. The value of KI is

(a) −a2 − 2a − 1
(b) a2 − 2a + 1
(c) a3 − 2a2 + a [Correct]
(d) −a3 − 2a2 − a
Solution:
C(s) Gc (s)G(s) KI
= =
R(s) 1 + Gc (s)G(s) s(s + 1)2 + KI

Given that s = −a is one of the root of the characteristic equation, hence

−a(−a + 1)2 + KI = 0 =⇒ KI = a(1 − a)2 = a3 − 2a2 + a

1
4. Determine the value of a that maximizes the integral gain. [Ans=0.33; Range = 0.28
to 0.38]
Solution:
dKI d2 KI
= 3a2 − 4a + 1; = 6a − 4
da da2
dKI 1
= 0 =⇒ 3a2 − 4a + 1 = 0 =⇒ a = 1 or a =
da 3
d2 KI d2 KI
|a=1 = 2 | 1 = −2
da2 da2 a= 3
1
Hence, a = 3 maximizes the integral gain.
5. Let one of the closed-loop pole be s = −a, where a is the solution of the previous problem. Let the
other poles be given as s = −p1 and s = −p2 where p2 > p1 . The value of p2 − p1 =
[Ans=1; Range = 0.98 to 1.02]
Solution:
4
KI = a(1 − a)2 = a3 − 2a2 + a =
27
1 1 4
s(s + 1)2 + 4/27 = (s + )(s + )(s + )
3 3 3
1
Hence, p1 = 3 and p2 = 43 , so p2 − p1 = 1.
6. What is the effect on the damping factor (ζ) and natural frequency (ωn ) in a control system of
second order when the proportional-derivative compensation of the form (1 + Ks) where K > 0 is
used?

(a) ωn increases and ζ decreases

(b) ωn decreases and ζ increases
(c) ωn decreases and ζ remains unchanged
(d) ωn remains unchanged and ζ decreases
(e) ωn remains unchanged and ζ increases [Correct]

Solution: (e)

ωn2 ωn2
G(s) = ; Torignal (s) =
s(s + 2ζωn ) s2 + 2ζωn s + ωn2
ωn2 G(s)Gc (s) (1 + Ks)ωn2
G(s) = ; Gc (s) = 1 + Ks Tnew (s) = = 2
s(s + 2ζωn ) 1 + G(s)Gc (s) s + (2ζωn + Kωn2 )s + ωn2
Comparing characteristic equation of Torignal (s) ans Tnew (s) , we see that ωn remains unchanged
and ζ increases.

Common data for Q7-Q8

A phase-lead network is given by

1 + 0.16s
Gc (s) = .
1 + 0.01s

2
 7. The frequency at which the maximum phase lead occurs is rad/sec. [Ans= 25; Range
24.5 to 25.5]
Solution:
1 + aT s
Gc (s) = .
1 + Ts
r r
1 1 1 1 1 1 100
ω= × = × = × = = 25
aT T 0.16 0.01 0.4 0.1 4
8. The maximum phase shift (in degrees) provided by the compensator is . [Ans= 61;
Range 60 to 62]
Solution:
0.16 a − 1
14

a= = 16; φm = arcsin = arcsin = 61.04◦
0.01 a+1 16
9. The addition of a zero to the closed loop system close to the origin on the LHP the
rise time of the system.
(a) increases
(b) decreases [Correct]
(c) have no effect on

Solution: (b)
As discussed in Lecture-1 that the addition of a zero to a closed loop system will give faster response
resulting in decrease of rise time.

Common data for Q10-Q13

Consider the unity-feedback system given below

R(s) + C(s)
Gc(s) G(s)
-

10. When Gc = 1, the closed-loop has the form of a standard second order system with ζ = 0.5 and
ωn = 4. The transfer function G(s) is
4
(a) s(s+2)
4
(b) s(s+4)
16
(c) s(s+4) [Correct]
16
(d) s(s+4)+16

Solution:
C(s) ωn2 16
= 2 2
= 2
R(s) s + 2ζωn s + ωn s + 4s + 16

3
 16
Gc (s)G(s) G(s) 16 2
=⇒ = = 2 = s +4s
1 + Gc (s)G(s) 1 + G(s) s + 4s + 16 1 + s216
+4s

16
=⇒ G(s) =
s(s + 4)

11. If Gc (s) = s + 3, the velocity error constant Kv is . [Ans=12; Range=11.98 to 12.02]

Solution:
16
Kv = lim sGc (s)G(s) = lim s(s + 3) = 12
s−→0 s−→0 s(s + 4)
1 s+4


12. If Gc (s) = 4 s+2 , which of the following statements are true when compared with Gc (s) = 1 ?
(a) overshoot increases
(b) overshoot remains constant [Correct]
(c) overshoot decreases
(d) settling time increases [Correct]
(e) settling time remains constant
(f) settling time decreases
Solution:
C(s) Gc (s)G(s) 4
= = 2
R(s) 1 + Gc (s)G(s) s + 2s + 4
From the closed-loop transfer function, we have

ωn = 2; ζ = 0.5

For a second order underdamped system, we have

− √ πζ 4
Overshoot = 100 e 1−ζ 2 ; ts =
ζωn

Since ζ is same overshoot remains same. Using the above formula, toriginal
s = 2 and tnew
s = 4, so
settling time increases.
1
13. If Gc (s) = s+a , the system is stable for
(a) a = −2
(b) a = 0.5
(c) a = 1 [Correct]
(d) a = 2 [Correct]
Solution:
C(s) 16 16
= = 3
R(s) s(s + 4)(s + a) + 16 s + (a + 4)s2 + 4as + 16
The Routh table is

s3 1 4a
s2 a+4 16
4a2 +16a−16
s1 a+4 0
s0 16

4
For stability

1. a + 4 > 0 =⇒ a > −4
2
4a + 16a − 16 √ √
2. > 0 =⇒ a < −2 − 2 2 or a > −2 + 2 2
a+4
√
Hence, a > −2 + 2 2 i.e a > 0.828 makes the system stable.