Anda di halaman 1dari 11

E1C06 11/03/2010 21:23:9 Page 326

326 Chapter 6 Stability

3. What would happen to a physical system that becomes unstable?


4. Why are marginally stable systems considered unstable under the BIBO
definition of stability?
5. Where do system poles have to be to ensure that a system is not unstable?
6. What does the Routh-Hurwitz criterion tell us?
7. Under what conditions would the Routh-Hurwitz criterion easily tell us the
actual location of the system’s closed-loop poles?
8. What causes a zero to show up only in the first column of the Routh table?
9. What causes an entire row of zeros to show up in the Routh table?
10. Why do we sometimes multiply a row of a Routh table by a positive constant?
11. Why do we not multiply a row of a Routh table by a negative constant?
12. If a Routh table has two sign changes above the even polynomial and five sign
changes below the even polynomial, how many right–half-plane poles does the
system have?
13. Does the presence of an entire row of zeros always mean that the system has jv
poles?
14. If a seventh-order system has a row of zeros at the s3 row and two sign changes
below the s4 row, how many jv poles does the system have?
15. Is it true that the eigenvalues of the system matrix are the same as the closed-
loop poles?
16. How do we find the eigenvalues?

Problems
1. Tell how many roots of the following polynomial are Determinehow many closed-loop poles liein theright
in the right half-plane, in the left half-plane, and on half-plane, in the left half-plane, and on the jv-axis.
the jv-axis: [Section: 6.2] 5. How many poles are in the right half-plane, in the
PðsÞ ¼ s þ 3s þ 5s þ 4s þ s þ 3
5 4 3 2 left half-plane, and on the jv-axis for the open-loop
system of Figure P6.1? [Section: 6.3]
2. Tell how many roots of the following polynomial are
in the right half-plane, in the left half-plane, and on
R(s) s2 + 4s – 3 C(s)
the jv-axis: [Section: 6.3]
s4 + 4s3 + 8s2 + 20s +15
PðsÞ ¼ s5 þ 6s3 þ 5s2 þ 8s þ 20
FIGURE P6.1
3. Using the Routh table, tell how many 6. How many poles are in the right half-plane, the left
poles of the following function are in half-plane, and on the jv-axis for the open-loop
the right half-plane, in the left half- system of Figure P6.2? [Section: 6.3]
plane, and on the jv-axis: [Section: 6.3]
sþ8 R(s) –6 C(s)
TðsÞ ¼
s5  s4 þ 4s3  4s2 þ 3s  2 s6 + s5 – 6s4 + s2 + s – 6

4. The closed-loop transfer function of a system is FIGURE P6.2


[Section: 6.3]
7. Use MATLAB to find the pole
s3 þ 2s2 þ 7s þ 21 locations for the system of
TðsÞ ¼ 5
s  2s4 þ 3s3  6s2 þ 2s  4 Problem 6.
E1C06 11/03/2010 21:23:9 Page 327

Problems 327

8. Use MATLAB and the Symbolic 15. Given the unity feedback system of Figure P6.3 with
Math Toolbox to generate a
8
Routh table to solve Problem 3. GðsÞ ¼
sðs6  2s5  s4 þ 2s3 þ 4s2  8s  4Þ
9. Determine whether the unity feedback
system of Figure P6.3 is stable if tell how many closed-loop poles are located in the
[Section: 6.2] right half-plane, in the left half-plane, and on the jv-
240 axis. [Section: 6.3]
GðsÞ ¼ 16. Repeat Problem 15 using MATLAB.
ðs þ 1Þðs þ 2Þðs þ 3Þðs þ 4Þ
17. Consider the following Routh table. Notice that the
R(s) + E(s) C(s) s5 row was originally all zeros. Tell how many roots
G(s) of the original polynomial were in the right half-

plane, in the left half-plane, and on the jv-axis.
[Section: 6.3]
FIGURE P6.3
10. Use MATLAB to find the pole
locations for the system of s7 1 2 1 2
Problem 9. s6 1 2 1 2
11. Consider the unity feedback system of Figure P6.3 s5 3 4 1 0
with s4 1 1 3 0
1 s3 7 8 0 0
GðsÞ ¼
4s2 ðs2 þ 1Þ s2 15 21 0 0
s1 9 0 0 0
Using the Routh-Hurwitz criterion, find the region
s0 21 0 0 0
of the s-plane where the poles of the closed-loop
system are located. [Section: 6.3]
12. In the system of Figure P6.3, let 18. For the system of Figure P6.4, tell how
many closed-loop poles are located in
Kðs þ 2Þ the right half-plane, in the left half-
GðsÞ ¼ plane, and on the jv-axis. Notice that
sðs  1Þðs þ 3Þ
there is positive feedback. [Section: 6.3]
Find the range of K for closed-loop stability.
[Section: 6.4] R(s) + E(s) 18 C(s)
13. Given the unity feedback system of Figure P6.3 with + s5 + s4 – 7s3 – 7s2 – 18s
[Section: 6.3]
84 FIGURE P6.4
GðsÞ ¼
sðs7 þ 5s6 þ 12s5 þ 25s4 þ 45s3 þ 50s2 þ 82s þ 60Þ
19. Using the Routh-Hurwitz criterion, tell how many
tell how many poles of the closed-loop transfer func- closed-loop poles of the system shown in Figure P6.5
tion lie in the right half-plane, in the left half-plane, lie in the left half-plane, in the right half-plane, and
and on the jv-axis. [Section: 6.3] on the jv-axis. [Section: 6.3]
14. Using the Routh-Hurwitz criterion and the unity
feedback system of Figure P6.3 with R(s) + 507 C(s)
s4+3s3+10s2+30s+169
1 –
GðsÞ ¼
2s4 þ 5s3 þ s2 þ 2s
1
s
tell whether or not the closed-loop system is stable.
[Section: 6.2] FIGURE P6.5
E1C06 11/03/2010 21:23:9 Page 328

328 Chapter 6 Stability

20. Determine if the unity feedback system of Figure 28. Find the range of gain, K, to ensure stability in the
P6.3 with unity feedback system of Figure P6.3 with [Section:
6.4]
Kðs2 þ 1Þ
GðsÞ ¼ Kðs  2Þðs þ 4Þðs þ 5Þ
ðs þ 1Þðs þ 2Þ GðsÞ ¼
ðs2 þ 12Þ
can be unstable. [Section: 6.4]
21. For the unity feedback system of Figure P6.3 with 29. Find the range of gain, K, to ensure stability in the
unity feedback system of Figure P6.3 with [Section:
Kðs þ 6Þ 6.4]
GðsÞ ¼
sðs þ 1Þðs þ 4Þ Kðs þ 2Þ
GðsÞ ¼ 2
ðs þ 1Þðs þ 4Þðs  1Þ
determine the range of K to ensure stability.
[Section: 6.4] 30. Using the Routh-Hurwitz criterion, find the value of
22. In the system of Figure P6.3, let K that will yield oscillations for the unity feedback
system of Figure P6.3 with [Section: 6.4]
Kðs  aÞ
GðsÞ ¼ K
sðs  bÞ GðsÞ ¼
ðs þ 77Þðs þ 27Þðs þ 38Þ
Find the range of K for closed-loop stability when:
31. Use the Routh-Hurwitz criterion to find the range
[Section: 6.4]
of K for which the system of Figure P6.6 is stable.
a. a < 0; b<0 [Section: 6.4]
b. a < 0; b>0
c. a > 0; b<0 R(s) + E(s) C(s)
K(s2 – 2s + 2)
d. a > 0; b>0 –
23. For the unity feedback system of
Figure P6.3 with 1
s2 + 2s + 4
Kðs þ 3Þðs þ 5Þ
GðsÞ ¼ FIGURE P6.6
ðs  2Þðs  4Þ
32. Repeat Problem 31 for the system of
determine the range of K for stability. [Section: 6.4] Figure P6.7. [Section: 6.4]
24. Repeat Problem 23 using MATLAB.
R(s) + E(s) K(s + 2) C(s)
25. Use MATLAB and the Symbolic s(s + 1)(s + 3)

Math Toolbox to generate a
Routh table in terms of K to
solve Problem 23. s+6
s+7
26. Find the range of K for stability for the unity feed-
back system of Figure P6.3 with [Section: 6.4] FIGURE P6.7
33. Given the unity feedback system of Figure P6.3 with
Kðs þ 4Þðs  4Þ
GðsÞ ¼ Kðs þ 4Þ
ðs2 þ 3Þ GðsÞ ¼
sðs þ 1:2Þðs þ 2Þ
27. For the unity feedback system of Figure P6.3 with
find the following: [Section: 6.4]
Kðs þ 1Þ a. The range of K that keeps the system stable
GðsÞ ¼
s4 ðs þ 2Þ b. The value of K that makes the system oscillate
c. The frequency of oscillation when K is set to the
find the range of K for stability. [Section: 6.4] value that makes the system oscillate
E1C06 11/03/2010 21:23:9 Page 329

Problems 329

34. Repeat Problem 33 for [Section: 6.4] a. Find the range of K for stability.
Kðs  1Þðs  2Þ b. Find the frequency of oscillation when the system
GðsÞ ¼ is marginally stable.
ðs þ 2Þðs2 þ 2s þ 2Þ
41. Using the Routh-Hurwitz criterion and the
35. For the system shown in Figure P6.8, find the unity feedback system of Figure P6.3 with
value of gain, K, that will make the system oscil- [Section: 6.4]
late. Also, find the frequency of oscillation.
[Section: 6.4] K
GðsÞ ¼
sðs þ 1Þðs þ 2Þðs þ 5Þ
R(s) + + 1 C(s)
K s(s+7)(s+3)
– – a. Find the range of K for stability.
b. Find the value of K for marginal stability.
s c. Find the actual location of the closed-loop poles
when the system is marginally stable.
FIGURE P6.8 42. Find the range of K to keep the system shown in
Figure P6.9 stable. [Section: 6.4]
36. Given the unity feedback system of
Figure P6.3 with [Section: 6.4]
R(s) + E(s) K C(s)
Ksðs þ 2Þ s
GðsÞ ¼ –
ðs  4s þ 8Þðs þ 3Þ
2

s–1
a. Find the range of K for stability.
s2 + 2s + 1
b. Find the frequency of oscillation when the system
is marginally stable. FIGURE P6.9

37. Repeat Problem 36 using MATLAB.


43. Find the value of K in the system of
Figure P6.10 that will place the closed-
38. For the unity feedback system of Figure P6.3 with loop poles as shown. [Section: 6.4]
Kðs þ 2Þ
GðsÞ ¼
ðs2 þ 1Þðs þ 4Þðs  1Þ
R(s) + K C(s)
s2
find the range of K for which there will be only two – s2
closed-loop, right–half-plane poles. [Section: 6.4] +
2 +
39. For the unity feedback system of Figure P6.3 with s
[Section: 6.4]
K 1
GðsÞ ¼ s+1
ðs þ 1Þ3 ðs þ 4Þ

a. Find the range of K for stability.
b. Find the frequency of oscillation when the system
is marginally stable.
σ
40. Given the unity feedback system of Figure P6.3 with
[Section: 6.4]
K
GðsÞ ¼
ðs þ 49Þðs2 þ 4s þ 5Þ FIGURE P6.10 Closed-loop system with pole plot
E1C06 11/03/2010 21:23:10 Page 330

330 Chapter 6 Stability

44. The closed-loop transfer function of a system is 48. A linearized model of a torque-controlled crane
hoisting a load with a fixed rope length is
s2 þ K 1 s þ K 2
TðsÞ ¼
s4 þ K 1 s3 þ K 2 s2 þ 5s þ 1 X T ðsÞ 1 s2 þ v20
PðsÞ ¼ ¼
F T ðsÞ mT s2 ðs2 þ av20 Þ
Determine the range of K 1 in order for the system to
be stable. What is the relationship between K 1 and qffiffiffiffiffi
g
K 2 for stability? [Section: 6.4] where v0 ¼ L ; L ¼ the rope length, mT ¼ the mass
45. For the transfer function below, find the constraints of the car, a ¼ the combined rope and car mass, f T ¼
on K 1 and K 2 such that the function will have only the force input applied to the car, and xT ¼ the
two jv poles. [Section: 6.4] resulting rope displacement (Marttinen, 1990). If
the system is controlled in a feedback configuration
K1 s þ K2 by placing it in a loop as shown in Figure P6.11, with
TðsÞ ¼
s4 þ K 1 s3 þ s2 þ K2 s þ 1 K > 0; where will the closed-loop poles be located?

46. The transfer function relating the output engine fan


R(s) + C(s)
speed (rpm) to the input main burner fuel flow rate K P(s)
(lb/h) in a short takeoff and landing (STOL) fighter –
aircraft, ignoring the coupling between engine fan
speed and the pitch control command, is (Schier-
man, 1992) [Section: 6.4] FIGURE P6.11

1:3s7 þ 90:5s6 þ 1970s5 þ 15; 000s4 þ 3120s3  41;300s2  5000s  1840


GðsÞ ¼
s8 þ 103s7 þ 1180s6 þ 4040s5 þ 2150s4  8960s3  10; 600s2  1550s  415 49. The read/write head assembly arm of a computer
hard disk drive (HDD) can be modeled as a rigid
a. Find how many poles are in the right half-plane, rotating body with inertia I b : Its dynamics can be
in the left half-plane, and on the jv-axis. described with the transfer function
b. Is this open-loop system stable?
47. An interval polynomial is of the form XðsÞ 1
PðsÞ ¼ ¼
FðsÞ I b s2
PðsÞ ¼ a0 þ a1 s þ a2 s2 þ a3 s3 þ a4 s4 þ a5 s5 þ   

with its coefficients belonging to intervals where X(s) is the displacement of the read/write
xi  ai  yi , where xi, yi are prescribed constants. head and F(s) is the applied force (Yan, 2003).
Kharitonov’s theorem says that an interval polyno- Show that if the HDD is controlled in the con-
mial has all its roots in the left half-plane if each one figuration shown in Figure P6.11, the arm will
of the following four polynomials has its roots in the oscillate and cannot be positioned with any pre-
left half-plane (Minichelli, 1989): cision over a HDD track. Find the oscillation
frequency.
K1 ðsÞ ¼ x0 þ x1 s þ y2 s2 þ y3 s3 þ x4 s4 þ x5 s5 þ y6 s6 þ   
K2 ðsÞ ¼ x0 þ y1 s þ y2 s2 þ x3 s3 þ x4 s4 þ y5 s5 þ y6 s6 þ    50. A system is represented in state space as
K3 ðsÞ ¼ y0 þ x1 s þ x2 s2 þ y3 s3 þ y4 s4 þ x5 s5 þ x6 s6 þ    2 3 2 3
K4 ðsÞ ¼ y0 þ y1 s þ x2 s2 þ x3 s3 þ y4 s4 þ y5 s5 þ x6 s6 þ    0 1 3 0
x_ ¼ 4 2 2 4 5x þ 4 1 5u
Use Kharitonov’s theorem and the Routh-Hurwitz 1 4 3 0
criterion to find if the following polynomial has any y ¼ ½1 1 0 x
zeros in the right–half-plane.
Determine how many eigenvalues are in the right
PðsÞ ¼ a0 þ a1 s þ a2 s2 þ a3 s3 half-plane, in the left half-plane, and on the jv-axis.
2  a0  4; 1  a1  2; 4  a2  6; a3 ¼ 1 [Section: 6.5]
E1C06 11/03/2010 21:23:10 Page 331

Problems 331

51. Use MATLAB to find the eigenval- of gain, K, that will keep the system stable. Can the
ues of the following system: system ever be unstable for positive values of K?
2 3
2 3 56. A common application of control systems is in
0 1 0 0 regulating the temperature of a chemical process
6 7 6 7
x_ ¼ 4 0 1 4 5x þ 4 0 5u (Figure P6.13). The flow of a chemical reactant to a
process is controlled by an actuator and valve. The
1 1 8 0
reactant causes the temperature in the vat to
y ¼ ½0 0 1 x change. This temperature is sensed and compared
to a desired set-point temperature in a closed loop,
where the flow of reactant is adjusted to yield the
52. The following system in state space
desired temperature. In Chapter 9, we will learn
represents the forward path of a unity
how a PID controller is used to improve the per-
feedback system. Use the Routh-
formance of such process control systems. Figure
Hurwitz criterion to determine if
P6.13 shows the control system prior to the addition
the closed-loop system is stable. [Sec-
of the PID controller. The PID controller is
tion: 6.5]
replaced by the shaded box with a gain of unity.
2 3 2 3 For this system, prior to the design of the PID
0 1 0 0 controller, find the range of amplifier gain, K, to
x_ ¼ 4 0 1 3 5x þ 4 0 5u keep the system stable.
3 4 5 1
y ¼ ½0 1 1 x
Actuator Chemical
Desired Future PID and heat
53. Repeat Problem 52 using temperature controller Amplifier valve process
Actual
MATLAB. set point + 1 0.7 temperature
1 K
– s + 0.4 s2 + 1.7s + 0.25
54. A Butterworth polynomial is of the form
 2n 0.1
n s s + 0.1
Bn ðsÞ ¼ 1 þ ð1Þ ; n > 0:
vc Temperature
sensor

Use the Routh-Hurwitz criteria to find the zeros of FIGURE P6.13 Block diagram of a chemical process control
a Butterworth polynomial for: system

a. n ¼ 1; 57. A robot arm called ISAC (Intelligent Soft Arm


b. n ¼ 2 Control) can be used as part of a system to feed
people with disabilities (see Figure P6.14(a)). The
control system guides the spoon to the food and
DESIGN PROBLEMS then to a position near the person’s mouth. The arm
55. A model for an airplane’s pitch loop is uses a special pneumatically controlled actuator
shown in Figure P6.12. Find the range called a rubbertuator. Rubbertuators consist of rub-
ber tubes covered with fiber cord. The actuator
contracts in length when pneumatic pressure is
increased and expands in length when pressure is
Commanded
Controller Aircraft dynamics decreased. This expansion and contraction in length
pitch angle + K(s + 1) s + 10 Pitch angle can drive a pulley or other device. A video camera
(s + 4.85) s2 + 0.6s + 9 provides the sight for the robot and the tracking

loop (Kara, 1992). Assume the simplified block
1
diagram shown in Figure P6.14(b) for regulating
the spoon at a distance from the mouth. Find the
Gyro range of K for stability. (Use of a program with
FIGURE P6.12 Aircraft pitch loop model symbolic capability is recommended.)
E1C06 11/03/2010 21:23:10 Page 332

332 Chapter 6 Stability

Desired Actual
force Controller Plant force
R(s) + 63 × 106 C(s)
K
(s+30)(s+140)(s + 2.5)

FIGURE P6.16 Cutting force control system (Reprinted with


permission of ASME.)

factors that affect cutting force are time varying


and not easily predicted. However, assuming
the simplified force control model shown in Figure
P6.16, use the Routh-Hurwitz criterion to find the
range of K to keep the system stable (Rober, 1997).
60. Transportation systems that use magnetic levitation
can reach very high speeds, since contact friction at
(a) the rails is eliminated (see Figure P6.17(a)). Electro-
magnets can produce the force to elevate the vehi-
Desired Actual cle. Figure P6.17(b) is a simulation model of a
spoon Rubbertuator spoon control system that can be used to regulate the
position Controller and load position
R(s) +
magnetic gap. In the figure, Zvin ðsÞ represents a
K(s+0.01)(s+6) 10 C(s)
s(s+20)(s+100) s2+10s+29
voltage proportional to the desired amount of levi-

tation, or gap. Zvout ðsÞ represents a voltage propor-
tional to the actual amount of levitation. The plant
(b)
models the dynamic response of the vehicle to
FIGURE P6.14 a. ISAC used for feeding (Courtesy of
Kazuhiko Kawamura, Vanderbilt University.)
b. simplified block diagram

58. Often an aircraft is required to tow


another vehicle, such as a practice
target or glider. To stabilize the
towed vehicle and prevent it from
rolling, pitching, and yawing, an autopilot is built
into the towed vehicle. Assume the block diagram
shown in Figure P6.15 represents the autopilot roll
control system (Cochran, 1992). Find the range of K
to keep the roll angle stable.

Commanded Roll Actual


roll angle Compensator Actuator dynamics roll angle
φc(s) + K(s+0.6)(s+6) 200 500 φ (s)

– (s+0.1)(s+100) s2+12s+100 s(s+6)


(a)

FIGURE P6.15 Towed vehicle roll control Controller Plant


Zvin(s) + 7570 Zvout(s)
K(s+0.8)(s+103)
59. Cutting forces should be kept constant during – s (s+62.61)(s–62.61)
machining operations to prevent changes in spindle
speeds or work position. Such changes would dete- (b)
riorate the accuracy of the work’s dimensions. A FIGURE P6.17 a. A magnetic levitation transportation system
control system is proposed to control the cutting (# Japan Air Lines/Photo Researchers); b. simplified block
force. The plant is difficult to model, since the diagram (# 1998 IEEE)
E1C06 11/03/2010 21:23:10 Page 333

Problems 333

signals from the controller (Bittar, 1998). Use the The circuit will oscillate if it is designed to have
Routh-Hurwitz criterion to find the range of gain, poles on the jv-axis.
K, to keep the closed loop system stable. a. Show that the transfer function for the passive
61. A transfer function from indoor radiator power, network in the circuit is given by
_
QðsÞ; to room temperature, T(s), in an 11 m2
V 2 ðsÞ 1
room is ¼  
V 1 ðsÞ 1 1 2 2
TðsÞ 1  106 s2 þ 1:314  109 s þ 2:66  1013
1þ 2þ 3
PðsÞ ¼ ¼ sRC sRC sRC
_
QðsÞ s3 þ 0:00163s2 þ 5:272  107 s þ 3:538  1011
b. Show that the oscillator’s characteristic equation
where Q _ is in watts and T is in  C. (Thomas, 2005). is given by
1
The room’s temperature will be controlled by 1K   ¼ 0;
embedding it in a closed loop, such as that of Figure 1 1 2 2
1þ 2þ 3
P6.11. Find the range of K for closed-loop stability. sRC sRC sRC
62. During vertical spindle surface grinding, adjust-
R2
ments are made on a multi-axis computer numerical where K ¼
R1
control (CNC) machine by measuring the applied
force with a dynamometer and applying appropriate c. Use the Routh-Hurwitz criterion to obtain the
corrections. This feedback force control results in oscillation condition and the oscillation
higher homogeneity and better tolerances in the frequency.
resulting finished product. In a specific experiment 64. In order to obtain a low-cost lithium-ion battery
with an extremely high feed rate, the transfer func- charger, the feedback loop of Figure P6.3 is used,
tion from the desired depth of cut (DOC) to applied where GðsÞ ¼ Gc ðsÞPðsÞ. The following transfer
force was functions have been derived for G(s) (Tsang,
2009):
FðsÞ KC
¼
DOCðsÞ
1þ 2
KC

KC 1 R1 R2 C1 C2 s2 þ ðR1 C1 þ R2 C1 þ R2 C2 Þs þ 1
ms þ bs þ k K f Ts þ 1 PðsÞ ¼
C1 ð1 þ R2 C2 Þs

where k ¼ 2:1104 N/m; b ¼ 0:78 Ns/m; m ¼ 1:2 KI


Gc ðsÞ ¼ K p þ
104 Kg; K C ¼ 1:5  104 N/mm and T ¼ 0:004 s: Kf s
is a parameter that is varied to adjust the system.
If R1 ¼ 0:15 V; R2 ¼ 0:44 V; C1 ¼ 7200 F; and C2 ¼
Find the range of K f under which the system is
170 F use the Routh-Hurwitz criteria to find the
stable (Hekman, 1999).
range of positive KP and KI for which the system is
63. Figure P6.18 depicts the schematic diagram of a closed-loop stable.
phase shift oscillator.
65. Figure P6.19 is a simplified and linearized block
diagram of a cascade control system, which is used
to control water level in a steam generator of a
nuclear power plant (Wang, 2009,).
R1 R2 In this system, the level controller, GLC(s), is the
master controller and the feed-water flow controller,
GFC(s) is the slave controller. Using mass balance
C C C equations, the water level would ordinarily be
+
– regarded as a simple integration process of water
+ + flow. In a steam generator, however, steam flow rate
V1(t) R R R V2(t) and the cooling effect of feed-water change the
– – dynamics of that process. Taking the latter into
account and ignoring the much-less pronounced
impact of changes in steam flow rate, a first-order
FIGURE P6.18 Phase shift oscillator lag plus time delay is introduced into the transfer
E1C06 11/03/2010 21:23:12 Page 334

334 Chapter 6 Stability

Controlled
Set point level
R(s) + X (s) + Y (s) Qw (s) C (s)
GLC (s) GFC (s) GV (s) Gfw (s)
_ _

FIGURE P6.19

2 3 2 3
function, Gfw(s), relating the controlled level, C(s), V_ 2 3 V
6 7 11:7 6:8 61:6K 7:7K 6 7
to feed-water flow rate, Qw(s) as follows: 6 r_ 7 6  3:5 6r 7
6 7¼6 24 66:9K 8:4K 7
76 7
6_ 7 4 0 1 0 0 56 c 7
CðsÞ K 1 et1 S 2e2S 4c 5 4 5
Gf w ðsÞ ¼ ¼ ¼ 1 0 10 0
Qw ðsÞ sðT 1 s þ 1Þ sð25s þ 1Þ Y_ g Yg
2 In this model, V ¼ bicycle’s lateral velocity, r ¼

sð25s þ 1Þð2s2 þ 2s þ 1Þ bicycle’s yaw velocity, c ¼ bicycle’s yaw accelera-
tion, and Y g ¼ bicycle’s center of gravity coordinate
where K1 ¼ 2 is the process gain, t 1 ¼ 2 is the pure on the y-axis. K is a controller parameter to be
time delay, and T 1 ¼ 25 is the steam generator’s chosen by the designer (Ozg€€ uner, 1995). Use the
time constant. (The expression et1 s represents Routh-Hurwitz citerion to find the range of K for
a time delay. This function can be represented by which the system is closed-loop stable.
what is known as a Pade approximation. This ap-
proximation can take on many increasingly compli- PROGRESSIVE ANALYSIS AND DESIGN PROBLEMS
cated forms, depending upon the degree of accuracy 67. High-speed rail pantograph. Problem 21 in Chapter 1
required. Here we use the Pade approximation, discusses active control of a pantograph mechanism
1 for high-speed rail systems. In Problem 79(a), Chap-
ex , and specific numerical values for ter 5, you found the block diagram for the active
x2
1þxþ pantograph control system. Using your solution for
2! Problem 79(a) in Chapter 5 and the Routh-Hurwitz
the considered steam generator.)
criterion, find the range of controller gain, K, that will
The dynamic characteristics of the control valve
keep the system stable (O’Connor, 1997).
are approximated by the transfer function:
Q ðsÞ Kv 1
Gv ðsÞ ¼ w ¼ ¼ , where Kv is the 68. Control of HIV/AIDS. The HIV infection linear-
YðsÞ T v s þ 1 3s þ 1 ized model developed in Problem 82, Chapter 4, can
valve gain and Tv is its time constant. be shown to have the transfer function
Given that: GFC ðsÞ ¼ K PFC þ K DFC s ¼ 0:5 þ 2s
YðsÞ 520s  10:3844
and GLC ðsÞ ¼ K PLC þ K DLC s ¼ 0:5 þ K s, use the PðsÞ ¼ ¼
Routh-Hurwitz criterion to find the range of the U 1 ðsÞ s3 þ 2:6817s2 þ 0:11s þ 0:0126
level controller’s derivative gain, K DLC ¼ K > 0, Desired virus
that will keep the system stable. count change + U1(s) Virus count change, Y(s)
G(s) P(s)
66. Look-ahead information can be used to automati- –
cally steer a bicycle in a closed-loop configuration.
A line is drawn in the middle of the lane to be
followed, and an arbitrary point is chosen in the FIGURE P6.20
vehicle’s longitudinal axis. A look-ahead offset is
calculated by measuring the distance between the It is desired to develop a policy for drug delivery to
look-ahead point and the reference line and is used maintain the virus count at prescribed levels. For
by the system to correct the vehicle’s trajectory. A the purpose of obtaining an appropriate u1 ðtÞ;
linearized model of a particular bicycle traveling on feedback will be used as shown in Figure P6.20
a straight-line path at a fixed longitudinal speed is (Craig, 2004).
E1C06 11/03/2010 21:23:12 Page 335

Cyber Exploration Laboratory 335

0.6154
Torque
controller Armature Motive TL (s) Motor
Ref. Speed Speed & power resistance Armature torque angular Vehicle
error amplifier Ra(s)
signal controller
UC (s) Ua(s) current T(s) _ speed speed
Ev(s) Ia(s)
Rv(s) K p s + 40 + 10 s + 6 + + 1 Ω (s) 0.3 V(s)
1 1.8
s s 7.226s 4.875
+ _ _ _ _
Feedback Friction
Feedback Eb(s) torque
current signal Back emf
speed signal Tf (s)
KCS Ia(s)
KSS Ω (s) 0.1

2
Current sensor
sensitivity

0.5
Speed sensor
sensitivity

0.0443

FIGURE P6.21

As a first approach, consider GðsÞ ¼ K; a constant have been substituted. It is assumed here that the
to be selected. Use the Routh-Hurwitz criteria to speed controller has a proportional gain, Kp, to be
find the range of K for which the system is closed- adjusted. Use the Routh-Hurwitz stability method
loop stable. to find the range of positive Kp for which the system
69. Hybrid vehicle. Figure P6.21 shows the HEV system is closed-loop stable (Graebe, 1995).
presented in Chapter 5, where parameter values

Cyber Exploration Laboratory


Experiment 6.1
Objectives To verify the effect of pole location upon stability. To verify the effect
upon stability of loop gain in a negative feedback system. R(s) + C(s)
G(s)
Minimum Required Software Packages MATLAB, Simulink, and the –
Control System Toolbox
H(s)
Prelab
1. Find the equivalent transfer function of the negative feedback system of Figure FIGURE P6.22
P6.22 if
K
GðsÞ ¼ and HðsÞ ¼ 1
sðs þ 2Þ2
2. For the system of Prelab 1, find two values of gain that will yield closed-loop,
overdamped, second-order poles. Repeat for underdamped poles.
3. For the system of Prelab 1, find the value of gain, K, that will make the system
critically damped.
E1C06 11/03/2010 21:23:12 Page 336

336 Chapter 6 Stability

4. For the system of Prelab 1, find the value of gain, K, that will make the system
marginally stable. Also, find the frequency of oscillation at that value of K that
makes the system marginally stable.
5. For each of Prelab 2 through 4, plot on one graph the pole locations for each case
and write the corresponding value of gain, K, at each pole.

Lab
1. Using Simulink, set up the negative feedback system of Prelab 1. Plot the step
response of the system at each value of gain calculated to yield overdamped,
underdamped, critically damped, and marginally stable responses.
2. Plot the step responses for two values of gain, K, above that calculated to yield
marginal stability.
3. At the output of the negative feedback system, cascade the transfer function
1
G1 ðsÞ ¼
s2 þ 4
Set the gain, K, at a value below that calculated for marginal stability and plot the
step response. Repeat for K calculated to yield marginal stability.

Postlab
1. From your plots, discuss the conditions that lead to unstable responses.
2. Discuss the effect of gain upon the nature of the step response of a closed-loop
system.

Experiment 6.2
Objective To use the LabVIEW Control Design and Simulation Module for
stability analysis.

Minimum Required Software Package LabVIEW with the Control Design


and Simulation Module

Prelab Select six transfer functions of various orders and use Routh-Hurwitz to
determine their stability.

Lab Create a LabVIEW VI that receives the order and the coefficients of the
characteristic equation and outputs the location of the poles and information
regarding stability.

Postlab Verify the stability of the systems from your Prelab.

Bibliography
Ballard, R. D. The Riddle of the Lusitania. National Geographic, April 1994, National
Geographic Society, Washington, D.C., 1994, pp. 68–85.