Lecture 23
Lecture 23
Control Systems
Lecture 23
Final Examination Review
and Practice Exercises
1/42
MECE 3350 - C. Rossa 1 / 42 Lecture 23
Gain and phase - review
For a generic transfer function G(s)
Qn
(s + zi )
G(s) = k Qmi=1
k=1
(s + pk )
we can evaluate the phase at a frequency ω by letting s = jω.
The phase is
n m
X X
∠G(jω) = ∠|k| + ∠(jω + zi ) − ∠(jω + pk )
i=1 k=1
2/42
MECE 3350 - C. Rossa 2 / 42 Lecture 23
Bode plot building blocks
1 - Constant gain
→ Gain: |k| or 20 log(|k|)
→ Phase: φ = 0 ∀ ω if k > 0, −180◦ otherwise
frequency [rad/s]
-90
0
-20
-2 -1.5 -1 -0.5 0 0.5 1 -2 -1.5 -1 -0.5 0 0.5 1
frequency [rad/s] frequency [rad/s]
3/42
MECE 3350 - C. Rossa 3 / 42 Lecture 23
Bode plot building blocks
0
90
-40
-2 -1.5 -1 -0.5 0 0.5 1 -2 -1.5 -1 -0.5 0 0.5 1
frequency [rad/s] frequency [rad/s]
0
0
Magnitude (dB)
-5
-15 -45
-20
-25 -90
4/42
MECE 3350 - C. Rossa 4 / 42 Lecture 23
Bode plot building blocks
5 - Real zero: G(s) = s
ω0
+ 1, ω0 ∈ <∗
25
90
Magnitude (dB)
20
Phase (deg)
15
10 45
0 0
Magnitude (dB)
0
-10 -40
180 0
Phase (deg)
Phase (deg)
90 -90
0 -180
10-1 100 101 10-1 100 101
Frequency (rad/s) Frequency (rad/s)
5/42
MECE 3350 - C. Rossa 5 / 42 Lecture 23
Open loop vs closed loop stability
Open-loop stability
T (s) = C (s)G(s) (1)
→ Evaluate the location of the poles of C (s)G(s)
Closed-loop stability
C (s)G(s)
T (s) =
1 + C (s)G(s)
→ Evaluate the location of the zeros of 1 + C (s)G(s)
s+a
Example: If C (s)G(s) = s+b
6/42
MECE 3350 - C. Rossa 6 / 42 Lecture 23
Cauchy’s argument principle
7/42
MECE 3350 - C. Rossa 7 / 42 Lecture 23
The Nyquist Stability Criterion
A open-loop transfer function L(s) is closed-loop stable if and only if the
number of counterclockwise encirclements of the −1 + 0j point is equal to the
number of poles of L(s) with positive real parts
Z =N +P
Nyquist plot
→ The contour at infinity maps to a single point
→ ω = 0 (starting point)
→ω→∞
→ Point where the plot crosses the real and imaginary axis
8/42
MECE 3350 - C. Rossa 8 / 42 Lecture 23
Gain and phase margins
1 + C (s)G(s) = 0
If |C (s)G(s)| = 1 and ∠|C (s)G(s)| = ±180◦ , the characteristic equations is
zero
50
1 0
Imaginary Axis
Magnitude (dB)
0 -100
-90
-135
-1
-1 Real Axis 0 Phase (deg)
-180
10-2 10-1 100 101 102
Frequency (rad/s)
9/42
MECE 3350 - C. Rossa 9 / 42 Lecture 23
Gain and phase margins
0 0
Magnitude (dB)
-180 -180
Phase (deg)
10/42
MECE 3350 - C. Rossa 10 / 42 Lecture 23
Phase and gain margin
Phase margin
Step 2 - Find the phase of G(jω) at ωc for ωc found in Step 1, i.e. ∠G(jωc )
Gain margin
11/42
MECE 3350 - C. Rossa 11 / 42 Lecture 23
State space model - back to temporal domain
State of a system: The set of variables that provides the future state and
output of the system for a given input.
12/42
MECE 3350 - C. Rossa 12 / 42 Lecture 23
Practice problems
Practice Exercises
13/42
MECE 3350 - C. Rossa 13 / 42 Lecture 23
Exercise 133
1
G(s) =
s + 10
by hand for ω = 1, 2, 5, 10, 20, 50 and 100 rad/s. Then, sketch the Bode
plot of G(s)and compare the results. The Bode plot can be obtained in Matlab
1
1
Gain: 0.095, 0.0981, 0.0894, 0.0707, 0.0447, 0.0196, 0.0099
Phase: -5.71, -11.3, -26.6, -45, -63.4, -78.7, -84.3 14/42
MECE 3350 - C. Rossa 14 / 42 Lecture 23
Exercise 134
Sketch the Bode plot for the following open-loop transfer function and estimate
the stability margins. After completing the and sketches, verify your results
using Matlab.
2000
L(s) =
s(s + 200)
Matlab script
bode(tf([2000],[1 200 0]))
15/42
MECE 3350 - C. Rossa 15 / 42 Lecture 23
Exercise 135
Sketch the Bode plot for the following open-loop transfer function and estimate
the stability margins. After completing the and sketches, verify your results
using Matlab.
s +2
L(s) =
s(s + 1)(s + 5)(s + 10)
Matlab script:
s = tf([1 0],[1]);
bode((s+2)/(s*(s+1)*(s+5)*(s+10)))
16/42
MECE 3350 - C. Rossa 16 / 42 Lecture 23
Exercise 136
Sketch the Bode plot for the following open-loop transfer function and estimate
the stability margins. After completing the and sketches, verify your results
using Matlab.
1
L(s) =
s 2 (s + 10)
Matlab script:
s = tf([1 0],[1]);
bode(1/(s*s*(s+10)))
17/42
MECE 3350 - C. Rossa 17 / 42 Lecture 23
Bode Diagram
50
0
Magnitude (dB)
-50
-100
-150
-200
-180
Phase (deg)
-210
-240
-270
10-1 100 101 102 103
Frequency (rad/s)
18/42
MECE 3350 - C. Rossa 18 / 42 Lecture 23
Exercise 137
Sketch the Bode plot for the following open-loop transfer function and estimate
the stability margins. After completing the and sketches, verify your results
using Matlab.
s +2
L(s) =
s(s + 10)(s 2 + 2s + 2)
Matlab script:
s = tf([1 0],[1]);
bode((s+2)/(s*(s+10)*(s*s+2*s+2)))
19/42
MECE 3350 - C. Rossa 19 / 42 Lecture 23
Exercise 138
Draw the Nyquist plot for the system shown. Using the Nyquist stability
criterion, determine the range of k for which the system is stable.
20/42
MECE 3350 - C. Rossa 20 / 42 Lecture 23
Exercise 138 - continued
Answer
-0.4
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
Real Axis
For positive k, note that the magnitude of the Nyquist plot as it crosses the
negative real axis is 0.1, hence k < 10 for stability.
21/42
MECE 3350 - C. Rossa 21 / 42 Lecture 23
Exercise 139
The Nyquist plot for a control system resembles the one shown below. What is
the phase margin(s)?2
2
−20 log(α), +20 log(β) 22/42
MECE 3350 - C. Rossa 22 / 42 Lecture 23
Exercise 140
Determine the range of k for which the following system is stable by making a
Bode plot for k = 1 and imagining the magnitude plot sliding up or down until
instability results.
k(s + 3)
G(s) =
s + 30
3
Stable ∀ k > 0 23/42
MECE 3350 - C. Rossa 23 / 42 Lecture 23
Exercise 141
Determine the range of k for which the following system is stable by making a
Bode plot for k = 1 and imagining the magnitude plot sliding up or down until
instability results.
k
G(s) =
(s + 10)(s + 1)2
4
Stable for k < 242 24/42
MECE 3350 - C. Rossa 24 / 42 Lecture 23
Exercise 142
The Bode plot of an unknown circuit has been obtained experimentally. Sketch
the Nyquist plot of the system based on the Bode plot.
0
Magnitude (dB)
-5
-10
-15
40
Phase (deg)
30
20
10
0
10-1 100 101 102 103
Frequency (rad/s)
25/42
MECE 3350 - C. Rossa 25 / 42 Lecture 23
Exercise 142 - continued
Answer
0.4
0.2
Imaginary Axis
-0.2
-0.4
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Real Axis
26/42
MECE 3350 - C. Rossa 26 / 42 Lecture 23
Exercise 143
5
k = 7.81 27/42
MECE 3350 - C. Rossa 27 / 42 Lecture 23
Exercise 144
A two thank system is controlled by a motor adjusting the input valve and
ultimately varying the output flow rate. The system has the transfer function
Q(s) 1
= P(s) = 3
I(s) s + 10s 2 + 29s + 20
28/42
MECE 3350 - C. Rossa 28 / 42 Lecture 23
Exercise 144 - solution
h 0 1 0
i h 0
i
ẋ = 0 0 1 x + 0 p
−20 −29 −10 1
y = [ 1 0 0 ]x
29/42
MECE 3350 - C. Rossa 29 / 42 Lecture 23
Exercise 145
An automatic depth control system for a robot submarine is shown in the
figure. The depth is measured by a pressure transducer. The gain of the stern
place actuator is k = 1 when the vertical velocity is 25 m/s. Determine a state
variable representation of the system.
30/42
MECE 3350 - C. Rossa 30 / 42 Lecture 23
Exercise 145 - continued
h 0 1 0
i h 0
i
ẋ = 0 0 1 x + 0 r
−1/3 −5/3 −5/3 1/3
y = [ 1 2 1 ]x
31/42
MECE 3350 - C. Rossa 31 / 42 Lecture 23
Exercise 146
32/42
MECE 3350 - C. Rossa 32 / 42 Lecture 23
Exercise 146 - solution
" 0 1 0 0
# " #
0
− k − b k 0 1
m1 m1 m1 m1
ẋ = x + u
0 0 0 1 0
k 0 − k − b 1
m2 m2 m2
y = [ 0 0 1 0 ]x
33/42
MECE 3350 - C. Rossa 33 / 42 Lecture 23
Exercise 147
34/42
MECE 3350 - C. Rossa 34 / 42 Lecture 23
Exercise 147 - solution
0 1
0
ẋ = x + r
−25 −6 25
y = [ 1 0 ]x
35/42
MECE 3350 - C. Rossa 35 / 42 Lecture 23
Exercise 148 - From 2018 final exam
Radio telescopes are used to study the radio frequency portion of the
electromagnetic spectrum emitted by astronomical objects. One critical
problem of driving large telescopes is the form of the system transfer function
that has a structural resonance. A large telescope with a diameter of 20
meters, for example, is subject to large wind gust torques that can affect
positioning accuracy. Consider an antenna, drive motor, and amplifier system
with the following transfer function
200
G(s) =
(s + 1)(s 2 + 5s + 100)
36/42
MECE 3350 - C. Rossa 36 / 42 Lecture 23
Exercise 149 - From 2018 deferred final exam
Radio telescopes are used to study the radio frequency portion of the
electromagnetic spectrum emitted by astronomical objects. One critical
problem of driving large telescopes is the form of the system transfer function
that has a structural resonance. A large telescope with a diameter of 20
meters, for example, is subject to large wind gust torques that can affect
positioning accuracy. Consider an antenna, drive motor, and amplifier system
with the following transfer function
s
G(s) = 150
(s + 0.7)2 (s 2 + s + 49)
37/42
MECE 3350 - C. Rossa 37 / 42 Lecture 23
Exercise 150 - From 2018 final exam
The Nyquist plot below was obtained for the open-loop transfer function kL(s)
for a given positive value of k.
Nyquist Diagram
0.4
0.2
Imaginary Axis
-0.2
-0.4
-1 0
Real Axis
If L(s) has one unstable pole, which of the following root-locus best
approximates the root-locus of the closed-loop feedback system as k → ∞?6
Imaginary Axis
Imaginary Axis
Imaginary Axis
Imaginary Axis
0 0 0 0
Root Locus
6
Answer: (b), why? 38/42
MECE 3350 - C. Rossa 38 / 42 Lecture 23
Research opportunities
bio
MECHA
TRONICS LAB
www.biomechatronics.ca
39/42
MECE 3350 - C. Rossa 39 / 42 Lecture 23
The end
40/42
MECE 3350 - C. Rossa 40 / 42 Lecture 23