UNIVERSITY OF RUHUNA
Faculty of Engineering
End-Semester 7 Examination in Engineering: November 2022
Module Number: EE6302 Module Name: Control System Design
Model Answers
Q3. a) i) A graphical representation of the closed-loop poles on the S-plane as a
system parameter is varied.
[2.0 Marks]
ii) The root locus can be used to describe qualitatively the performance of a
system as various parameters are changed. For example, the effect of gain
on percent overshoot, settling time, and peak time. Besides transient
response, the root locus also gives a graphical representation of a system’s
stability.
[1.5 Mark]
iii)
The zeros of the closed-loop transfer function, 𝐺𝑂𝐿 (𝑠) :
𝐾 𝑁(𝑠) = 0
The zeros of open-loop transfer function 𝐺𝐶𝐿 (𝑠).
𝐾 𝑁(𝑠) = 0
∴ the zeros of the closed-loop transfer function equal to the zeros of open-
loop transfer function.
[1.5 Mark]
b) i) Open-loop poles = −15, −2.5 ± 0.866𝑗
There are no open-loop zeros.
1
𝐺(𝑠) =
(𝑠 + 15)(𝑠 + 2.5 − 0.866𝑗)(𝑠 + 2.5 + 0.866𝑗)
1
𝐺(𝑠) =
(𝑠 + 15)((𝑠 + 2.5)2 − (0.866𝑗)2 )
1
𝐺(𝑠) =
(𝑠 + 15)(𝑠 2 + 5𝑠 + 6.25 + 0.75)
1
𝐺(𝑠) =
(𝑠 + 15)(𝑠 2 + 5𝑠 + 7)
[1.5 Mark]
Page 1 of 10
ii) Properties of the root locus
Open-loop poles = −15, −2.5 ± 0.866𝑗
There are no open-loop zeros.
𝐿. 𝐻. 𝑆 = −∠(𝑠 + 15) − ∠(𝑠 + 2.5 + 0.866𝑗) − ∠(𝑠 + 2.5 + 0.866𝑗)
Substitute 𝑠 = −2 + 3.75𝑗
𝐿. 𝐻. 𝑆 = −∠(−2 + 3.75𝑗 + 15) − ∠(−2 + 3.75𝑗 + 2.5 + 0.866𝑗) − ∠(−2
+ 3.75𝑗 + 2.5 − 0.866𝑗)
𝐿. 𝐻. 𝑆 = −∠(13 + 3.75𝑗) − ∠(0.5 + 4.616𝑗) − ∠(0.5 + 2.884𝑗)
𝐿. 𝐻. 𝑆 = −16.09𝑜 − 83.82𝑜 − 80.160
𝐿. 𝐻. 𝑆 = −180.07𝑜
is satisfied. Therefore, 𝑠 = −2 + 3.75𝑗 is on the root locus.
[1.0 Mark]
iii) Root locus: 1 + 𝐾𝐺(𝑠) = 0
1
1+𝐾 =0
(𝑠 + 15)(𝑠 2 + 5𝑠 + 7)
(𝑠 + 15)(𝑠 2 + 5𝑠 + 7) + 𝐾 = 0
𝑠 3 + 20𝑠 2 + 82 𝑠 + 105 + 𝐾 = 0
On the imaginary axis, 𝑠 = 𝑗𝜔
1 + 𝐾𝐺(𝑗𝜔) = 0
(𝑗𝜔)3 + 20(𝑗𝜔)2 + 82 (𝑗𝜔) + 105 + 𝐾 = 0
−𝑗𝜔3 − 20𝜔2 + 82𝑗𝜔 + 105 + 𝐾 = 0
−20𝜔2 + 105 + 𝐾 + 𝑗𝜔(−𝜔2 + 82) = 0
−𝜔2 + 82 = 0
𝜔 = ±9.05
[1.5 Marks]
iv)
Consider the angle of departure at open-loop pole 𝑠 = −2.5 + 0.866𝑗 is 𝜃𝑝 .
0.866
0 − 𝜃𝑝 − 900 − tan−1 = 180𝑜
15 − 2.5
𝜃𝑝 = −900 − 3.96𝑜 − 180𝑜
𝜃𝑝 = −273.96𝑜 = 86.04𝑜
Root locus is symmetrical about real axis. Therefore, angle of departure at
open-loop pole 𝑠 = −2.5 − 0.866𝑗 is −86.04𝑜 .
[1.0 Mark]
Page 2 of 10
v) Gain K on the imaginary axis crossings of the root locus
−20𝜔2 + 105 + 𝐾 = 0
𝐾 = 20𝜔2 − 105
𝐾 = 20 × 9.052 − 105 = 1533.05
The closed-loop system is stable for 𝐾 < 1533.05
[1.0 Mark]
vi)
[1.0 Mark]
Q4. a) i) Proportional
𝑦𝑐 (𝑡) = 𝐾𝑝 𝑒(𝑡)
Integral
𝑡
𝑦𝑐 (𝑡) = 𝐾𝑖 ∫ 𝑒(𝑡) 𝑑𝑡
0
Derivative
𝑑𝑒(𝑡)
𝑦𝑐 (𝑡) = 𝐾𝑑
𝑑𝑡
[1.5 Marks]
ii) The relationship between the damping ratio (𝜁) and the percent overshoot is given
by the following equation.
−𝜁𝜋
√1−𝜁2
𝑀𝑝 % = 100 𝑒
From this equation damping ratio corresponding to certain overshoot can be
calculated.
sin 𝜃 = 𝜁
In the s-plane, root locus and the sin 𝜃 = 𝜁 line be drawn as follows. The gain
corresponding to the crossing point of the root locus and sin 𝜃 = 𝜁 line results the
desired percent overshoot.
Page 3 of 10
[1.5 Marks]
iii) When the gain is increased, the steady state error is reduced.
The slow response of an overdamped system can be made faster by
increasing the gain.
However, the maximum overshoot will increase with increasing gain.
[1.0 Mark]
b) i) −𝜁𝜋
√1−𝜁2
𝑀𝑝 % = 100 𝑒
𝑀𝑝 % = 20
−𝜁𝜋
√1−𝜁2
20 = 100 𝑒
−𝜁𝜋
√1−𝜁2
𝑒 = 0.2
−𝜁𝜋
= ln (0.2)
√1 − 𝜁 2
−𝜁𝜋 = −1.609√1 − 𝜁 2
(𝜁𝜋)2 = 1.6092 (1 − 𝜁 2 )
𝜁 = 0.456
[1.0 Mark]
Page 4 of 10
Peak time of the compensated system; 𝑇𝑝 = 0.5 s.
Suppose the second order dominant poles of the compensated system are
𝑠 = −𝜎 + 𝑗𝜔
𝜋
𝑇𝑝 =
𝜔
𝜋
𝜔= = 6.283
𝑇𝑝
sin 𝜃 = 0.456
𝜎
tan 𝜃 =
𝜔
𝜎 = 𝜔 tan 𝜃 = 6.283 tan(sin−1 0.456) = 3.219
𝑠 = −3.219 + 𝑗6.283
[1.0 Mark]
Root locus of the compensated system;
1 + 𝐾𝐺𝑃𝐷 (𝑠)𝐺(𝑠) = 0
𝐺𝑃𝐷 (𝑠) = 𝑠 + 𝑧𝑐
1
1 + 𝐾(𝑠 + 𝑧𝑐 ) =0
(𝑠 + 1)(𝑠 + 2)(𝑠 + 12)
Using the second property of the root locus
∑ ∠𝜃𝑧 − ∑ ∠𝜃𝑝 = 180(2ℎ ± 1)
∠(𝑠 + 𝑧𝑐 ) − (∠(𝑠 + 1) + ∠(𝑠 + 2) + (𝑠 + 12)) = 180
𝑠 = −3.219 + 𝑗6.283 should be a point on the root locus of the compensated
system.
𝜃𝑧𝑐 − (∠(−3.219 + 𝑗6.283 + 1) + ∠(−3.219 + 𝑗6.283 + 2)
+ (−3.219 + 𝑗6.283 + 12)) = 180
6.283 6.283 6.283
𝜃𝑧𝑐 − tan−1 − tan−1 − tan−1 = 180
−3.219 + 1 −3.219 + 2 −3.219 + 12
𝜃𝑧𝑐 = 66.02
6.283
𝜃𝑧𝑐 = 66.02 = tan−1
−3.219 + 𝑧𝑐
𝑧𝑐 = 6.015
[2.0 Marks]
Page 5 of 10
Now we need to design the compensator gain.
𝑠 = −3.219 + 𝑗6.283 is a point on the root locus of the compensated system.
1
1 + 𝐾(𝑠 + 𝑧𝑐 ) =0
(𝑠 + 1)(𝑠 + 2)(𝑠 + 12)
1
1 + 𝐾(𝑠 + 6.015) =0
(𝑠 + 1)(𝑠 + 2)(𝑠 + 12)
1
1 + 𝐾(−3.219 + 𝑗6.283 + 6.015) =0
(−3.219 + 𝑗6.283 + 1)(−3.219 + 𝑗6.283 + 2)(−3.219 + 𝑗6.283 + 12)
(−3.219 + 𝑗6.283 + 1)(−3.219 + 𝑗6.283 + 2)(−3.219 + 𝑗6.283 + 12)
𝐾=−
(−3.219 + 𝑗6.283 + 6.015)
𝐾 = 66.9
PD compensator = 66.9 (s+6.015)
[1.0 Mark]
Justify the second order approximation.
Closed-loop poles of the compensated system
1
1 + 66.9(𝑠 + 6.015) =0
(𝑠 + 1)(𝑠 + 2)(𝑠 + 12)
𝑠 3 + 15𝑠 2 + 104.9𝑠 + 426.404 = 0
𝑠 = −8.565, −3.217 ± 𝑗6.279
Closed-loop zeros of the compensated system
𝑠 = −6.015
Closed-loop pole −8.565 and closed-loop zero −6.015 located close by.
Therefore, those can be canceled out. Hence second order approximation is
valid.
[1.0 Mark]
ii) 1
𝑒𝑠𝑠 = lim
𝑠→0 1 + 𝐺𝑂𝐿 (𝑠)
1
𝑒𝑠𝑠 = lim 1
𝑠→0 1 + 66.9(𝑠 + 6.015) (𝑠+1)(𝑠+2)(𝑠+12)
1
𝑒𝑠𝑠 = 1
1 + 66.9 × (0 + 6.015) (0+1)(0+2)(0+12)
𝑒𝑠𝑠 = 0.056
[1.0 Mark]
iii) Add a PI controller in addition to the PD controller designed above.
[1.0 Mark]
Page 6 of 10
Q5. a) 145𝑗𝜔
𝐺(𝑗𝜔) =
(𝑗𝜔 + 5)(𝑗𝜔 + 10)(𝑗𝜔 + 25)
20𝑗𝜔
𝐺(𝑗𝜔) = 𝜔 𝜔 𝜔
5(𝑗 5 + 1)10(𝑗 10 + 1)25(𝑗 25 + 1)
0.116𝑗𝜔
𝐺(𝑗𝜔) = 𝜔 𝜔 𝜔
(𝑗 5 + 1)(𝑗 10 + 1)(𝑗 25 + 1)
1 1
|𝐺(𝑗𝜔)| 𝑑𝐵 = 20 log 0.116 + 20 log |𝑗𝜔| + 20 log | 𝜔 | + 20 log | 𝜔 |
𝑗5 +1 𝑗 10 + 1
1
+ 20 log | 𝜔 |
𝑗 25 + 1
1 1 1
∠𝐺(𝑗𝜔) = ∠𝑗𝜔 + ∠ 𝜔 +∠ 𝜔 +∠ 𝜔
𝑗5 +1 𝑗 10 + 1 𝑗 25 + 1
Page 7 of 10
[3.0 Marks]
Page 8 of 10
b) i) 1 1
𝑒𝑠𝑠 = lim =
𝑠→0 1 + 𝐺𝑂𝐿 (𝑠) 1 + 𝐺𝑂𝐿 (0)
𝐾
𝐺𝑂𝐿 (𝑠) = 2
(𝑠 + 3𝑠 + 7)
𝐾
𝐺𝑂𝐿 (0) =
7
1
0.01 = 𝐾
1+ 7
𝐾 = 693
[1.0 Mark]
ii) −𝜁𝜋
√1−𝜁2
𝑀𝑝 % = 100 𝑒
𝑀𝑝 % = 20
𝜁 = 0.456
2𝜁
𝜙𝑃𝑀,𝑟𝑒𝑞 = tan−1 = 48.2𝑜
√−2𝜁 2 + √1 + 4𝜁 4
[1.0 Mark]
iii) Calculate the phase margin of the gain adjusted system
693
𝐺(𝑠) = 2
(𝑠 + 3𝑠 + 7)
693 693 7 − 𝜔2 + 3𝑗𝜔
𝐺(𝑗𝜔) = = ×
((𝑗𝜔)2 + 3𝑗𝜔 + 7) 7 − 𝜔 2 + 3𝑗𝜔 7 − 𝜔 2 − 3𝑗𝜔
693(7 − 𝜔2 + 3𝑗𝜔)
𝐺(𝑗𝜔) =
(7 − 𝜔 2 )2 + 9𝜔 2
1⁄
2 2 2
7 − 𝜔2 3𝜔
|𝐺(𝑗𝜔)| = 693 {( ) + ( ) }
(7 − 𝜔 2 )2 + 9𝜔 2 (7 − 𝜔 2 )2 + 9𝜔 2
At phase margin frequency |𝐺(𝑗𝜔)| = 1
1⁄
2 2 2
7 − 𝜔2 3𝜔
693 {( ) +( ) } =1
(7 − 𝜔 2 )2 + 9𝜔 2 (7 − 𝜔 2 )2 + 9𝜔 2
6932 [(7 − 𝜔2 )2 + 9𝜔2 ] = ((7 − 𝜔2 )2 + 9𝜔2 )2
𝜔2 = 𝑥
6932 = (7 − 𝑥)2 + 9𝑥
𝑥 = 695.47, −690.47
𝜔 = 26.37
Phase margin of the gain adjusted system
693
𝐺(𝑗𝜔) = | = 1∠ − 173.4
7 − 𝜔 2 + 3𝑗𝜔 𝜔=26.37
𝜙𝑃𝑀,𝑜𝑟𝑔 = 180 − 173.4 = 6.6
[2.0 Marks]
Page 9 of 10
Phase contribution required from the lead compensator
𝜙𝑙𝑒𝑎𝑑 = 𝜙𝑃𝑀,𝑟𝑒𝑞 − 𝜙𝑃𝑀,𝑜𝑟𝑔 + 10 = 48.2 − 6.6 + 10 = 51.6
1−𝛽
𝜙𝑙𝑒𝑎𝑑 = 𝜙𝑙𝑒𝑎𝑑,𝑚𝑎𝑥 = sin−1
1+𝛽
1−𝛽
51.6 = sin−1
1+𝛽
𝛽 = 0.124
[1.5 Marks]
Compensator’s magnitude at the peak of the phase curve
1 1
𝐺𝑙𝑒𝑎𝑑 (𝑗𝜔𝑚𝑎𝑥 ) = = = 2.839 = 9.06 dB
√𝛽 √0.124
[1.0 Mark]
Frequency at which the uncompensated system’s magnitude equals to the
negative of the lead compensator’s magnitude at the peak of the
compensator’s phase curve.
|𝐺(𝑗𝜔)| = −9.06 dB = 0.353
6932 = 0.3532 [(7 − 𝑥)2 + 9𝑥]
3854047 = (7 − 𝑥)2 + 9𝑥
𝑥 = 1965.7, −1960.7
𝜔 = 44.3
[1.0 Mark]
𝜔 = 𝜔𝑚𝑎𝑥 = 44.3
1
𝜔𝑚𝑎𝑥 =
𝑇√𝛽
1 1
𝑇= = = 0.064
𝜔𝑚𝑎𝑥 √𝛽 44.3√0.124
[1.0 Mark]
1 1
1 𝑠+𝑇 1 𝑠 + 0.064 𝑠 + 15.625
𝐺𝑙𝑒𝑎𝑑 (𝑠) = = = 8.06
𝛽𝑠+ 1 0.124 𝑠 + 1 𝑠 + 126
𝛽𝑇 0.124×0.064
[0.5 Mark]
Page 10 of 10