Vous êtes sur la page 1sur 19

Answers to Selected Problems

CHAPTER 2
2.1.

(a) Loop equations: Let R R2 R3 and R 0 R1/R






1
1
et R1 LD
i  LD
i
CD 1
CD 2




1
1
0  LD
i1 R LD
i
CD
CD 2
(b) Node equations:
Node va:
 
 


1
1
1
1
1

vc
e
va  CDvb 
R1 R2 LD
R2
R1
Node vb:
CDva


1
CD vb 0
LD

Node vc:
 


1
1
1


va
v 0
R2
R2 R3 c

Copyright 2003 Marcel Dekker, Inc.

796

Linear Control System Analysis and Design

(c) State equations: Let x1 y1 i/CD, x2 iL, x2 DiL, ic CDy1 Cx1,


i1 i2 ic iL
2
3
1
0
6
7
C
7
A6
6
0
4 1
5
R
 0

L
R 1L
2

3
0
5
1
b 4
0
R 1L
T

cT 1 0

(d ) Note: Utilizing Eqs. (1), (2), (8), and (9), obtain


x2 C x_ 1

x_ 2 C x 2

GD
2.2.

y1 t
1

0
2
ut R 1LCD R 0 CD R 0 1

(a) Loop equations:






1
1
E R1
i 
i
C1 D 1
C1 D 2



1
1
1

0
i R2 LD
i
C1 D 1
C1 D C2 D 2

10

(b) Node 1 between R1 and R2, node 2 between R2 and L, and node 3
between L and C2


 
 
1
1
1
1
v1 
v2
E
Node 1:
C1 D
R1
R2
R2
R1





 
1
1
1
1

v1
v2 
v 0
R2
R2 LD
LD 3

Node 2:

Node 3:




1
1
C2 D v3 0

v
LD 2
LD


Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

797

(c) Let x1 v1 vc1, x2 v3 vc2, x3 iL


3
3
2
2
1
1
1
0


6 R1 C1 7
6 R 1 C1
C1 7
7
7
6
6
7
7
6
6
1
7
7
6
6
b6
A6 0
0
7
7
0
C2 7
7
6
6
5
5
4
4 1
1
R2


0
L
L
L
2.3.

2.6.

Circuit 2: (b)
2
R1 R2
6
L
x_ 6
4
1
C

3
2 3
1
1
7
7
6
L
L7
6 7
5x 4 5u
0
0

(c)
GD

D4

BD3

KD2

K2
BKa D Ka Kb  K 2

where Ka K1 K2, Kb K2 K3, K Ka Kb


(d ) Let
x2 x_ 1

x1 xa
x4 x_ 3
2

2.10.

x3 xb

u f t yt xb
0

6 K
6
A6
4 0

K2
0

0
0

K2

Kb

(a) LD Ri e

07
7
7
15

1

ki i f 

2 3
0
607
6 7
c6 7
415

2 3
0
617
6 7
b6 7
405

0
xa xb

l1
l2

M2 D2 B2 D K2 xb f2
M1 D2 B1 D K1 xa

l2
f f ki i
l1 2

Let
l2
a
l1

M1
aM2 b;
a

bD2 cD dxb ki i

Copyright 2003 Marcel Dekker, Inc.

B1
aB2 c
a

and

K1
aK2 d
a

798

Linear Control System Analysis and Design

(b) There are only three independent states:


x1 xb  x2 x_ b  x3 i
x_ 1 x2
d
c
K
x_ 2  x1  x2 1 x3
b
b
b
R
l
x_ 3 x3 u
L
L
2.11.
2.12.

and

ue

Traction force MD2x BDx


(b) T(t) (B1D)ya  (B1D)yb
0 (B1D)ya [JD2 (B2  B1)D K1]yb
(c) Let J J1 J2, y1 x1 ya, y2 x2 yb,
3
2 3
2
1
0
0
1
7
6B 7
6
7
6 17
60
0
1
7x 6 0 7u
x_ 6
7
6
6 7
4
4 1 5
K B1  B2 5
0
J
J
J
y1
JD2 B2 D K
d

u B1 DJD2 B2  B1 D K

x3 x2, and u T

y

1 0
0 1

y2
1

2
u JD B2  B1 D K

2.18. a

2.21.


C1 ad
C b
D y1 1 x1
a bc d
ab

KT v1(t) JLaD3ym (JRa BLa)D2ym (BRa KTKb)ym

CHAPTER 3
3.2.

yb t 2 2t A1 em1 t A2 em2 t
where
A1 2:01015

Copyright 2003 Marcel Dekker, Inc.

A2 0:01015


0
x
0

Answers to Selected Problems

799

Therefore
yb t 2 2t 2:010105e1:071t  0:01015e15:06t
3.5.

Check at t 0: yb(0)  2 2.01015  0.01015 0


(b) ct 0:36 0:2t A1 em1 t A2 em2 t A3 em3 t
Utilizing the initial conditions yields
p

A2 2=10 171:9 : A3
A1 0:5

3.7.

(a) Physical variables: x1 vc and x2 iL.




 4
104 106
10
A
b
0:1 465
0

cT 1 0

(b) vc t 48:95 A1 em1 t A2 em2 t where Dvc m1 A1 em1 t m2 A2 em2 t


Inserting the initial conditions yields
0 48:95 A1 A2
A1 1:15
3.8.

0:5  106 475:5A1  9989:5A2 


A2 50:1

(b) x(t) 3 1.5e3t  4.5et


c xt 0:18 0:2353t e0:5t 0:18 cos 2t  0:16 sin 2t
0:18 0:2353t 0:242e0:5t sin 2t  132

3.13.

"
(t
2

3 16
.

3.17.

e2t 2e4t

2e2t  2e4t

e2t e4t

2e2t  e4t
3

1
9 4t
2t
6 2  3e 2e 7
7
xt 6
43
9 4t 5
2t
 3e e
4
4
1
9
yt x1  3e2t e4t
2
2


e2t
a xth (tx0
2e2t


1  2et 2e2t
b xt
2et
(1) (a) 1 1, 2,3 1
j2

(t Z1 et Z2 et e jt Z3 et ejt

Copyright 2003 Marcel Dekker, Inc.

800

Linear Control System Analysis and Design

CHAPTER 4
4.1.
4.2.
4.4.

(a) f (t) et[t 0.333 sin (3t 180)]


(b) f (t) 1.3846  e5t 1.9612e2t sin (3t 191.3)
(c) x(t) 1  0.9898et 0.01701e2t sin (3t 216.9)
(b) With zero initial conditions, the Laplace transform yields:
Xb s

12
A
B

ss 10s 1:2874s 223:71 s s 10
C
D

s 1:2874 s 223:71

where
A 4:167  103

B 0:06445  103

C 4:81  103

D 1:128  106

xb t 4:167  103 0:6445  104 e10t  4:81  103 e1:2874t


 1:128  106 e223:7t
1
7
5
0:5


s 2 6s 4s 1 6s 3
1
9
28
30

h F s
s s 3 s 3  j1 s 3 j1

4.5.

d F s

4.6.

b iL t 0:4833e3:684t  0:5534 sin od t 61

4.7.

b s 2 2:8s 4X s

2s 2 4:3s 6:13
s

xt 3:0669 1:5e1:4t sin 1:416t  45:32


d s 5 4s 4 34s3 110s2 225s 250X s 4s 2  s  56
xt 10:837et sin 2t  85:233
0:99974 sin 5t 180 14:799e2t
4.8.

(a) Prob. 4.1(a): 0, Prob. 4.1(b): 18


13
(b) Prob. 4.2(a): 0, Prob. 4.2(c): 1

Copyright 2003 Marcel Dekker, Inc.

5
s2 25

Answers to Selected Problems

4 11
.

801

1 b xt 3 3:8348et sin t 237:53 0:2353e5t


c xt 0:30t  0:260 0:2712et sin t 102:53  0:0047e5t

4.12.

System (1)
a (s sIA1

s5
2

1
s2

1


1 s2
2


1
s5

 s2 7s 8 s 5:62s 1:438
"
#


X1 s
1
s 2 3s 1
Xs

 s 3 12s 2 35s 1
X2 s
Xs sI  A1 x0 sI  A1 bUs
(b) Gs CsB D

 
s2
1
0
1
0
Gs 01

2
s5 1
s5
s5
2

s 7s 8 s 5:62s 1:438
c Ys CXs 0
0:74836e

1 Xs X2 ! yt x2 t

1:4384t

 0:37336e5:5616t 0:6250

As a check: xi 0 Lim sXi s 1 and x2 0 Lim sX1 s


s!1

s!1

1sX2 s 1 and
4.15.

(1) s(A) { 1 j2, 1 j2, 1}

CHAPTER 5
55
.

5.11.

Gs

60s 3:8
s 10s 2 0:15s 8:0054

(2)  2,

 3,

(a) y [8 0
(b) y [4

Copyright 2003 Marcel Dekker, Inc.

 4,

0]x

22 36

18]x

 5

y1 0:6250:

802

Linear Control System Analysis and Design

5 12
.

a Gs

3
5
2t
3t
6  1  3e 2e 7
c xt 4 3 10
5
1  e3t
3

10
s 2s 3

CHAPTER 6
6.2.

(a) 0 < K < 1280/9; s1,2


j2.5827; for K 142.22
(c) 0 < K < 30; s1,2
j2.2356; for K 30

6.3.

(a) (s 2)(s 2.2


j2.2)

(c) (s 1
j1.414)(s 1
j 1.414)

( f ) (s 1)(s 2)(s  3)(s 5)


6.4. (a) None (c) None ( f ) 2
5
6.10. (a) For (1): Kp 1,Kv 13
,Ka 0, (b) For r1(t): (1) e(t)ss 13
5 (2) e(t)ss 0
6.12. (a) 20, 0, 0 (d) 1, 6, 0
6.13. (a): (a) 52 (b) 1 (c) 1
(d ): (a) 0 (b) 13 (c) 1
6.19. KGci ai KhKGci > 0 for a stable system and KGc0 a0 KhKGc0 for
e(t)ss 0
6.25. (b) K 4: (b) K 7: and K1 6.666 s1 e(t)ss 0.3;
6.27. (a) For K 100: K1 68.97 s1 and e(t)ss 0.0145; (b) for K 1000:
K1 689.7 s1 and e(t)ss 0.00145; (c) unstable for K 1000; stable
for K 100 and the corresponding FOM are Mp 1.345, tp 1.06 s,
and ts 3.286 s.

CHAPTER 7
7.2.

7.4.

(c) For K > 0, real-axis branches: none; four branches; g


45,

135 ; angles of departure:


45,
135 ; imaginary-axis crossing:

j2 with K 0. Branches are straight lines and coincide with the


asymptotes.
(e) Real-axis branch: 0 to 1; five branches; g
36,
108, 180 ;
s0  1.6; no breakaway or break-in points; angles of departure:

71.7,
33.8 ; and the imaginary-axis crossing:
j1 with K 30.
Locus crosses the
36 asymptotes.
(a) For K > 0, real-axis branches: 0 to  10; three branches; g
90 ;
s0 0; break-in point: 3.9113 (K 3.33095); breakaway point:
2.78698 (K 3.4388); and imaginary-axis crossing: none.
(b) K 0:05427

Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

803

(c) et 1:5951e2:2164t sin 2:1992t 207:274


(d )

o
0
0.2
0.4
0.6
0.8
1.0

a, deg

a, deg

1
0.9995
0.998
0.9951
0.9903
0.9832

0
6.1
12.3
18.5
24.9
31.3

1.4
1.8
2.0
2.4
2.8
3.2

0.9586
0.9153
0.8854
0.8103
0.7202
0.6252

44.7
58.5
65.5
79.3
92.3
104.1

Poles: p1 1 p2 7 p3 4 4


j3
(a) For K > 0, real-axis branch: 1 to  7; four branches; g
45,

135 ; angles of departure: 0, 180,


90 ; imaginary-axis crossing
j4 with K 1105; break-in and breakaway pints coincide at
s  4. For K < 0, real-axis branches: 1 to 1 and  7 to 1;
g
90, 180 ; angles of departure; 0,
90, 180 ; imaginary-axis
crossing: s 0 with K 175.
b  175 < K < 1105 c K 81
7.8. (1) (a) For K  50K1: three branches; g
90 for K1 <0; g 180
for K1 >0; s0  9.5; breakaway points:  7.72 for K > 0,  3.139 for
K < 0; break-in point: 5.362 for K < 0; imaginary-axis crossings:
at s 0 for K 50 and
j3.6993 for K 66.32.
(b) Stable for 66.32 < K < 50
7.11. (a) Real-axis branches: 3.2 to 1 (Note: The canceled pole and zero
at s 5 is a root of the characteristic equation and is not utilized
in determining the root locus for locating the remaining roots.)
6-branches: g
36,
108, 180, s0  22.5; Break-in point:
s 16.5; Angle of departures (given in order of closeness to the
imaginary axis in the 2nd quadrant): 144.4, 94.4, 4.2 ;
Imaginary-axis crossing:
j17.20 for K 1.869 107.
(b) KG KH 1:0735; p1 2
j17:21 p3 4 33:28
j28:05 p5
4:634 p6 44:46 p7 5
7.7.

7.12.

Cs
1:869  107 s 3:2

Rs s 4:634s 5s 44:46s
j17:21s 33:28
j28:05

1 a A 1:44

Copyright 2003 Marcel Dekker, Inc.

804

Linear Control System Analysis and Design

CHAPTER 8


(b) (1) G j0 1 90  G j1 0  270  ox oc 2:236,
G jox 0:2223 of 0:867


(2) G j0 1 90  G j1 0 360  ox oc 9:75
G jox 0:0952 of 0:4977


8.4. (1) b G j0 1 90  G j1 0 270  ox 35:4
G jox 0:05645
8.6. (a) It must have an initial slope of  40 dB/decade
(b) The phase-angle curve approaches an angle of 180 as o ! 0
8.9. Case II:

8.1.

Gs

8.10.
8.11.

a K0 31:7 b K2 2:56 oy 2
(a) Plant 1
Gs

8.13.
8.14.
8.15.
8.16.

2337:51 0:25s
s1 5s1 0:125s1 0:05s

0:4
ss 1
j4

(a) N  2, ZR 2, unstable (c) N 0, ZR 0, stable


(a) For stability K > 1.26759
(a) Stable for K > 2 (e) Stable for 18 < K < 90.12
(1) a K1 5:5654 of 4:845 b K1 3:345 of 3:15
c maximum K1 1:9429  105
(2) a K0 20:13 of 8:3 b K0 8:58 of 4:61
c Stable for 0 < K0 < 1

CHAPTER 9
9.1.
9.3.
9.5.
9.8.
9.11.

2 K1 7:94 Mm 1:326 om 7:81 of 6:94 oc 15:81


(4) K1 0:7075 Mm 1:318 om 0:658 of 0:6054 oc 1:508
Mm 1:30 K1 2:975 om 2:75
(1) (a) Mm 1 om 0 b Mm 1:326 om 5:4 c Mm 1:016
om 2:24
(4) a Mm 5:8904 om 3:82 c K0 0:18 for Mm 1:12
d om 0:39 g 56:16
Using a CAD program: (a) Mm 1.11293, om 0.435 (b) K 1.338
(c) a 6

Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

805

CHAPTER 10
10.3, 10.18.

System (1)

Dominant
roots

Other
roots

K1 s1

Tp,s

ts,s

Mo

Basic
Lag-compensated:
a 10, T 2

1.25
j 2.166

6.501

1.72

1.67

3.36

0.14

1.03
j 1.787

5.267
0.726

16.16

1.8

5.37

0.41

Lead-compensated:
a 0.1, T 0.25

2.23
j 3.86

3.123
40.42

3.13

0.916

1.78

0.10

Lag-lead-compensated:
a 10, T1 2, T2 0.25

2.08
j 3.6

0.598
2.897
40.40

30.18

0.967

4.15

0.244

Tachometer-compensated:
A 1.61, Kt 1, z 0.5
A 4.79, Kt 1, z 0.3

3.40
j 5.89
1.86
j 5.92

1.2
4.28

1.02
2.97

0.756

3.39
1.89

0.0
0.13

10.6.
10.7.

10.18.

(a) a 4, b 2.6653 (b) a 1.126 (c) lag compensator


(a) For the uncompensated system: from the performance specifications, for z 0.5, are Mp 1.16, tp 2 s, ts 4.39 s, K 2.933,
K1 1.76 s1, and the control ratio poles are  0.9235
j 1.601
and 5.153.
(b) jsj  4=Ts 2; assuming a 2, z 0.5, and jsj 2 yields
b 4.726
(c) A 5.21 and a 0.4232
(d) A lead compensator; Mp 1.59, tp 0.917 s, ts 2.02 s, and
K1 3.88 s1
(a) For z 0.5 the uncompensated system yields the following data:
K 25.01, K1 5 s1, Mp 1.163, tp 0.725 s, ts 1.615 s.
(b) Based upon the desired values of z and s 7, the desired
dominant poles are P1,2D  7
j12.124. One possible solution
is: let a 40, b 0.01, c 0.1; by applying the angle condition
and the use of a CAD package the following results are obtained:
A 6.0464, Kh 11.18, K1 126.1s1, Mp 1.161, tp 0.27 s,
ts 0.426 s.

Copyright 2003 Marcel Dekker, Inc.

806

Linear Control System Analysis and Design

10.20. See Prob.10.3 for basic system performance data. Based on the desired
performance specifications, the desired dominant poles are:
P1 2D 4
j6:928. Let Kx 1250 K1 be the gain term of G(s). By
applying method 1, the following results are obtained: K x 760.71,
Kt 0.5619, b 11.895, Mp 1.092, tp 0.725 s, ts 0.97 s, K1
2.369 s1. The other poles are: 6.296
j4.486 and 11. The plot of
c(t) has the desired simple second-order time response characteristics
due to the effect of the real root.

CHAPTER 11
11 1 a K 11935, of 3:10, g 51:16, Mp 1:161, tp 0:8654 s
ts 1:807 s K1 3:6387 s1 :
(b) 1=T 0:302 K 11935 A 7:7084 Mm 1:18
tp 1:104 s
ts 5:455 s
K1 28:05 s1 ;
Mp 1:203

of 2:53 g 51:1 :
Results for an Mm 1:16 are not as good as for Mm 1:18:
(c) T 0:25, K 213200, Mm 1:564, Mp 1:186, tp 0:3125 s
ts 1:273 s, K1 6:50 s1 ; for Mm 1:16: Kx 175800, Mp
1.076, tp 0:3372 s ts 1:105 s K1 5:36 s1  of 7:10,
g 66:46 :
(d) T1 3:311, T2 0:25, Kx 209000, Mm 1:564, Mp 1:212,
tp 0:3180 s, ts 2:857 s, K1 63:72 s1 ; for Mm 1:16: K
175800, Mp 1:104, tp 0:3442 s, ts 3:526 s K1 54:87 s1 
of 7:35 g 62:71 :


1 1 Ts
B
and  0:5
11 5 a Gs
 where T
8K 1 Ts
K
.

11 12 a At o 8 : G jo C jo=E jo  6:3 119  thus K1 250


of 8 and g  40 :
.

b Gc

s 10:5s 1
10s 10:05s 1

11.17. (a) K2x 2.02, ofx 2.15, Mm 1.398, Mp 1.29, tp 1.39 s, ts 3.03 s
(c) For g 50: u of T 8:6T 1:19 yields T 0.1384. From the
graphical construction, with o1 of =20 and o2 20 of 172,
obtain Kt 24.28 and A 180, thus
Cs
2907s 1s 7:225

Es s 2 s 1:108s 406:3

Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

807

The 0-dB crossing slope is 12 dB/oct at o 6, but the zero at


7.225 will have a moderating effect on the degree of stability.
(d ) Mm 1.47, Mp 1.30, tp 0.34 s, ts 1.03 s, K2 46.65; since the
values of Mp are essentially the same, a good improvement in
system performance has been achieved. g 51 and of 9:12.

CHAPTER 12
12 2
.

Gc s

2:31s 20
s 21

With this cascade compensator Mp 1.12, tp 1.493 s, ts 2.941 s,


K1 1.53. The specified MT (s) yields Mp 1.123, tp 1.493, ts 2.948,
K1 1.424.
12.4. As a first trial, use
Gc 

2:994s 2:1
s 2:223

With this cascade compensator Mp 1.03, tp 2.37 s, ts 2.94 s,


Kt 0.75. The specified MT (s) yields Mp 1.06, tp 2.2 s, ts 3.04 s,
K1 0.79.
12.8. (a) The pole at the origin of H(s) ensures c(t)ss 0; make pole b
nondominant: b 100
b KH 2500
(c) Chxaracteristic equation: s4 104s3 10,400s2 120,000s
320,000 0. Routh stability criterion reveals that all roots lie
in the left-half s plane. A plot of Lm GxH reveals that its 0-dB
axis crossing slope is 6 dB/oct, indicating a stable system.
The system poles are p1 4, p2  8.06, p3,4  45.97

j311.7.jy(tp)j 0.0025, tp 0.163 s. For jy(tp)j 0.01, K H 630,


which produces very little change in the value of tp.

CHAPTER 13
(1) (a) Not observable, (b) controllable, (c)Gs 2=s  2, (d) one
observable and two controllable states, (e) unstable
(4) (a) Observable, (b) controllable, (c) Gs s  1=s 1 s 2,
(d) Two observable and two controllable states, and (e) stable
13.5. (a) Add a cascade compensator Gc s 1=s 1
b Heq s k3 s2 2k3 k2 s k1

13.3.

Copyright 2003 Marcel Dekker, Inc.

808

Linear Control System Analysis and Design

(d) A desired control ratio is


MT s

Y s
A

Rs s2 4s 13s 25

e With A 162:5 kT 1 0:16923 0:08


g Geq s 325=ss 2 29s 113 K1 2:876
h Mp 1:12 tp 1:09 s ts 1:66 s
13.8. For system (1) (a)
M s

11700
s 12
j21s 20

k1 1

k2 0:091026

k3 0:003077

12t

13.12.

13.16.

(b) yt 1 1:025e
sin 21t 171:1  1:1584e20t
Mp 1:057 tp 0:215 s ts 0:273 s K1 11 s1
(a)
Y s
1:467s 1:5s 2

MT s
Rs s2 2s 2s 1:1s 2
s 1:5
 A 1:467
where Gc s
s1
Y s=Rs
As 1:5s 2
4
s 7 A0:5k4 k3 k2 s 3 11 A3k4 2:5k3 3:5k2 k1 s 2
5 A2:5k4 1:5k3 3k2 3:5k1 s 3Ak1
where k1 1 k2 0:8635 k3 2:43541 k4 0:5521
c Mp 1:015 tp 3:75 s ts 2:79 s K1 0:524 s1
(a)
MT s

195s 2
s 2s 2s 1:95s 100
2

(b) By applying the Guillemin-Truxal method, a reduced-order


compensator is obtained by dividing the denominator of Gc(s) by
its numerator to obtain Gc(s)  1/(s 99.9). The figures of merit
for MT(s) are Mp 1.042, tp 3.18 s, and ts 4.23 s, where for the
system implemented with the reduced-order compensator
Mp 1.056, tp 3.12 s, and ts 4.44 s.


T
13.19. k 8 2:5


5 3t 1 6t
1
3t
6t
x1 t e  e
x2 t 5e 4e
x0
1
3
3
13.20.

(1) (a) Completely controllable, (b) s ( Ap) {1, 2, 3},

Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

809

6 5 1
7
6
c T 4 3 2 1 5
0 2 1



d kT 42 33 6

CHAPTER 14
14.1.

m 1: (a)

For

SkT sjk15
14.6.

(a) p 21, KG 30.8


Gc

14.10.

s 4 7s 3 25s 2 39s
s 4 7s3 25s 2 54s 75
(b) No; need a compensator of the form

As a
s bs c

For system (2), (a) Three equilibrium points are


2 3
2 13
2 13
0
2
2
6 7
6
6
7
7
xb 4 0 5
xa 4 0 5
xc 4 0 5
0
2
2
(b) For xb,

2
3

3
0

6
J xb 4 2

0:5x3 7
5
2x1 1 x

CHAPTER 15
15.5.
15.7.
15.8.

a z 1 1:8z 2 2:19z 3 2:252z 4


a f 0 0; system is stable thus f 1 1:95122
a 1  2z 1 1:41z 2  0:41z 3 Cz z 1  0:2z 2 Ez
By use of Theorem 1:
ckT 2ck  1T   1:41ck  2T  0:41ck  3T 
ek  1T   0:2ek  2T 

15.11.

System 2
C  s G1 G2 sM  s

Copyright 2003 Marcel Dekker, Inc.

810

Linear Control System Analysis and Design

C  s
G1 G2 s

R s 1 G1 H  s G1 G2 s
Cz
G1 G2 z

Rz 1 G1 H z G1 G2 z
CHAPTER 16
16.1.

(1)
z-domain value

16.2.

16.3.

s Plane

Eq. (16.8)

Z esT

 1.5
3
 20
2
j3
10
j 25

0.94176
0.8868
0.4286
0.9167
j 0.1106
0.4201
j 0.5917

0.94176
0.88692
0.44933
0.91671
j 0.11054
0.39208
j 0616063

(2) zTU 0.2857, z 0.006738difference due to warping


(3) zTU 0.71598
j0.11834, z 0.20586
j0.69589difference due
to warping
(1)


Cs
Ks 8K
! 8K

Rs T s 2 10s 174s 10
10174 ! K 217:5
(b) For T 0.1!
j(0.6114/T )
j6.114 and (0.1)/T  1!
thus, the zero and poles of the desired control ratio do not lie in the
good Tustin region of Fig. 16.5. Thus:
0:6114
0:1
0:1
 7 ! T 0:08734 and
 10 ! T
0:01;
T
T
10
therefore,T 0.01.

16.5.
Dc z

Kc z  0:6703
z  0:06703

where

K 3:795  102 K1  0:2566 ! Kc 6:7615

Copyright 2003 Marcel Dekker, Inc.

Answers to Selected Problems

Mp 1:062 tp 0:4 s ts 0:6 s

811

and

K1 5:1577 s1

Cs
251:2s 0:005

Rs s 0:4061
j0:4747s 0:0005s 3:183s 200
Mp 1:078 tp 6:98 s ts 10:7 s K1 4:187 and A 10
16.10.

The remaining solutions are left to the instructor to obtain.


(a)
Gz z

4:967  105 Kx z 0:9934


! Kx 2:34105
z  1z  0:9802

Cz 1:163  104 z 0:9934

Rz
z  0:9900
j0:01152
(b) The PCT block diagram is similar to that of Fig. 16.14 d with
Dc s 1 and Gpa s=T 200=s 200 GA s:
200Kx
 K 2:34105
ss 2s 200 x
Cs
GPC s
466:21

Rs 1 GPC s s 0:99408
j1:159s 200:012

GPC s GA Gx s

The continuous-time system control ratio, for z 0.65, is:


Cs
Gx s
2:368

Rs 1 Gx s s 1
j1:170
(c) DIGTo reduce ts by about 1/2, select Dc s As 2=s 5;
Gs Dc sGp s 1 for z 0.65: K 200AK x 2874.6;
Kx 2.368, and A 6.0705.
Cs
K

;
Rs s 2:463
j2:881s 200:1
5:982z  0:9802
Dc zTU
z  0:9512
Cz
Dc zTU Gz z
7:0359  104 z 0:9934

Rz 1 Dc zTU Gz z
z  0:9752
j0:02810
DIRFor K x 2.368 and z 0.65,
z  0:09802 where A 1865:3.

Dc(z)

Cz
Dc zGz z
0:2194z 0:9934

Rz 1 Dc zGz z z  0:4393
j0:3507

Copyright 2003 Marcel Dekker, Inc.

A(z  0.9802)/

812

Linear Control System Analysis and Design

(d ) DIGFor K x 2.368 and z 0.65,


s 0:001 and A 9:384893:
Cs
Dc sGPC s

Rs 1 Dc sGPC s

Dc(s) = 0.1A(s + 0.01)/

Cs
468:4s 0:01

Rs s 0:98955
j1:1595s 0:01008s 200:01
Dc zTU

0:9385315z  0:9999
z  0:99999

Cz
Dc zTU Gz z

Rz 1 Dc zTU Gz z

1:1633  104 z 0:9934z  0:9999


z  0:990087
j0:01148126z  0:99989923

DIRFor Kx 2.368, z 0.65, and


K z  0:9999
! Kzc 0:986259 obtain:
Dc zTU zc
z  0:99999
Cz
Dc zGz z
1:1163104 z 0:9934z 0:9999

Rz 1Dc zGz z z 0:990087


j0:01145146z 0:99989923
(e) The tables summarize the results of the various designs presented
on the following page. The desired improvements using s-domain
compensation design rules (for both s- and z-domain analysis)
are obtained. Since the dominant poles and zeros lie in the
good Tustin region for the lag and lead units, there is a good
correlation between the DIG and DIR results.

System
Basic (a) Continuous
and (b)
DIG
DIR
(c) Lead
DIG

(d) Lag

DIR
DIG
DIR

Copyright 2003 Marcel Dekker, Inc.

Domain

Zeros
of C/R

s
z
s
z
z
s
z
z

Poles of C/R
1.000
j 1.170

0.9941
j 1.159, 200.012
0.9934
0.9900
j 0.01152

2.463
j 2.881, 200.1
0.9934
0.9752
j 0.02810
0.9934
0.4393
j 0.3507
0.01
 0.9895
j 1.160, 0.01008, 200.01
0.9934
0.990087
j 0.011481, 0.999899
0.9999
0.990087
j 0.001145146, 0.999899

Answers to Selected Problems


System
Basic (a) and (b)

(c) Lead

Continuous
DIG
DIR
DIG

(d) Lag

DIR
DIG

Domain

Kx

Mp

tp s

ts s

K1 s1

s
s
z
s
z
z
s
z
z

2.368
2.341
2.341
2.368

1.0682
1.0675
1.068
1.068
1.068
1.067
1.0767
1.077
1.076

2.69
2.72
2.70
1.096
1.09
0.05
2.71
2.71
2.72

3.90
3.94
3.93
1.59
1.58
0.07
4.15
4.16
4.17

1.184
1.165
1.171
2.875
2.874
2.84
11.71
11.71
11.68

DIR

16.13.

813

(a)
Gz z Gzo Gx z

2:073  107 z 3:695z 0:2653


z  1z  0:9900z  0:9704

Cz
2:073  107 z 3:695z 0:2653

Rz z  0:9959
j0:0047756z  0:9687
Mp 1:067 tp 6:94 s ts 9:92 s and K1 0:41604 s1
(b)
Gs GA sGx s

200Kx
ss 1s 3s 200

Cs
251:2

Rs s 0:4082
j0:4773s 3:184s 200
Mp 1:067 tp 6:94 s ts 9:90 s and K1 0:4187 s1
(c) PCT lead designDc s Ksc ss 1=s 3
Cs
1454

Rs s 0:8439
j0:9866s 4:312s 200
Mp 1:064 tp 3:47 s ts 4:88 s K1 0:808 s1  and
A = 5.7882
(d ) PCT lag designDc s A=10s 0:005=s 0:0005

Copyright 2003 Marcel Dekker, Inc.