ELEN90055 Control Systems

Worksheet 3 - Solutions to Starred Problems

Semester 2, 2021

1.
   
Y G1 G4
= G7 + G3 G6
R 1 + G1 G2 1 + G4 G5

2. (a) and (b): The transfer function is

Ae−sT
Y s(s+1) Ae−sT
= −sT = .
R 1 + Ae s2 + s + Ae−sT
s(s+1)

1− T2 s
Using the approximation e−sT ≈ 1+ T2 s
, the transfer function can be written as

1− T s
Y A 1+ T2 s A(1 − T2 s)
2
= = .
R s2 + s + A 1− T2 s (s2 + s)(1 + T2 s) + A(1 − T2 s)
1+ T s 2

So the characteristic equation is:

T T
(s2 + s)(1 + s) + A(1 − s) = 0
2 2
T 3 T 2 T
⇒ s + (1 + )s + (1 − A )s + A = 0
2 2 2

Routh’s Hurwitz array:

T
s3 2 1 − A T2 ⇒ T
2 >0 ⇒T >0
s2 1 + T2 A ⇒ T > −2
(1+ T2 )(1−A T2 )−A T2 2
s1 1+ T2
0 ⇒ 1 − AT4 + T2 (1 − A) > 0
s0 A 0 ⇒ A>0
2
So the conditions on A and T is that T > 0, A > 0 and 1 − AT4 + T2 (1 − A) > 0 (think why
we did not assume the first column of the Routh’s array is negative).

1
3. We first find the transfer function of the system:
K 1
Y (s) K
H(s) = = Js+kK s
K 1
=
R(s) 1 + Js+kK s
Js2 + kKs + K

K 1
⇒ Y (s) =
Js2 + kKs + K s
Using the initial value theorem we have

lim y(t) = lim sY (s) = 0.

t→0 s→∞

We now use the final value theorem to compute the steady state value of the output, but
before that we need to check the stability of the system. It can be easily verified that the
system is stable as J, kK and K are all positive. So we can apply the final value theorem:

lim y(t) = lim sY (s) = 1

t→∞ s→0

4. In this system, the equation for the error is

E(s) = R(s) −Y (s)

 G(s) 
= 1− R(s)
1 + G(s)
 1 
= R(s)
1 + G(s)
 s2 (s + 2) 
= 3 R(s)
s + 2s2 + 4s + 4
We now check the stability of the system to see if we can use the final value theorem.
Routh’s Hurwitz array:

s3 1 4
s2 2 4
2×4−1×4
s1 2 =2 0
s0 4

So all poles of the system are on the left-half plane and therefore the system is stable.

(a) R(s) = 1s :

s2 (s + 2)
lim e(t) = lim sE(s) = lim =0
t→∞ s→0 s→0 s3 + 2s2 + 4s + 4

2
 1
(b) R(s) = s2
:

s2 (s + 2) 1
lim e(t) = lim sE(s) = lim s =0
t→∞ s→0 s→0 s + 2s + 4s + 4 s2
3 2

2
(c) R(s) = s3
:

s2 (s + 2) 2
lim e(t) = lim sE(s) = lim s 3 =1
t→∞ s→0 s→0 s + 2s + 4s + 4 s3
2

5. (a)
D(s)
Y (s) 2 D(s) s+k
H1 (s) = = s +1
D(s)
= 2 = 2
R(s) 1 + 2 s + 1 + D(s) s + s + k + 1
s +1

1
Y (s) 2 1 1
H2 (s) = = s +1
D(s)
= 2 = 2
W (s) 1 + 2 s + 1 + D(s) s + s + k + 1
s +1

(b) The characteristic polynomial is s2 + s + k + 1 = 0. For the second order system, we

know all poles are on the laft-half plane if and only if all coefficients of the polyno-
mial are positive (or all negative). So k + 1 > 0, that is k > −1.
(c) Define e(t) = r(t) − y(t). Then

s2 + 1
   
s+k 1
E(s) = R(s) −Y (s) = 1 − 2
R(s) = 2
s +s+k+1 s +s+k+1 s

1
lim e(t) = lim sE(s) = (1)
t→∞ s→0 k+1
We now calculate the output from the disturbance w(t).
   
1 1 1
Y (s) = 2 W (s) = 2
s +s+k+1 s +s+k+1 s

1
lim y(t) = lim sY (s) = (2)
t→∞ s→0 k+1
From (1) and (2), we observe that we can make limt→∞ e(t) and limt→∞ y(t) arbitrarily
small by letting k to be large. Note that k should be greater than 0 to make the system
stable (see part (b)).