Farid Golnaraghi, Benjamin C. Kuo-Automatic Control Systems, 9th Edition - Solutions Manual (2009)

Ninth Edition
l
ua
an
M
ns
tio
lu
So
Farid Golnaraghi Benjamin C. Kuo

AutomaticControlSystems,9thEdition
A Chapter2Solution
ns Golnarraghi,Kuo
C
Chapter 2
2
21(a) 10;
Poless:s=0,0,1, (b) Poles:s=2,,2;
Zeross:s=2,,,. Zeros:s=0.
Thepoleandzeroats=1ccanceleachotther.
( Poles:s=0,1+j,1j;
(c) (d)Poles:s =0,1,2,.
Zeross:s=2.
2
2-2) a)
b)
c)
2
2-3)
M
MATLABcode
e:
21
AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo
clear all;
s = tf('s')
'Generated transfer function:'
Ga=10*(s+2)/(s^2*(s+1)*(s+10))
'Poles:'
pole(Ga)
'Zeros:'
zero(Ga)
Gb=10*s*(s+1)/((s+2)*(s^2+3*s+2))
'Poles:';
pole(Gb)
'Zeros:'
zero(Gb)
Gc=10*(s+2)/(s*(s^2+2*s+2))
'Poles:';
pole(Gc)
'Zeros:'
zero(Gc)
Gd=pade(exp(-2*s),1)/(10*s*(s+1)*(s+2))
'Poles:';
pole(Gd)
'Zeros:'
zero(Gd)
22
Polesandzerosoftheabovefunctions:
(a)
Poles:00101
Zeros:2
(b)
Poles:2.00002.00001.0000
Zeros:01
(c)
Poles:
1.0000+1.0000i
1.00001.0000i
Zeros:2
Generatedtransferfunction:
(d)usingfirstorderPadeapproximationforexponentialterm
Poles:
2.0000
1.0000+0.0000i
1.00000.0000i
Zeros:
23
2-4) Mathematical representation:
In all cases substitute and simplify. The use MATLAB to verify.
10 22 + 2 1 + 2 102 + 2
R= ;
2 ( 2 + 1)( 2 + 100)

10( j + 2)
1 = tan 1 2 +
2 2
( j + 1)( j + 10)
2
2
10( j + 2) ( j + 1)( j + 10) 22 + 2
=
( j + 1)( j + 10) ( j + 1)( j + 10)
2

10( j + 2)( j + 1)( j + 10)
2 = tan 1 1 +
2
a) =
( + 1)( + 100)
2 2 2 1
j + 2 j + 1 j + 10 1+ 2
=R
22 + 2 1 + 2 102 + 2
3 = tan 1 10 +
2 2
= R(e j1 e j2 e j3 )
10
102 + 2
= 1 + 2 + 3
10 1 + 2 9 + 2
R= ;
( 2 + 1) 2 ( 2 + 9)

10
1 1 +
2
( j + 1) ( j + 3)
2 1 = tan
1
10 ( j + 1)( j + 1)( j + 3) 1+ 2
=
( j + 1)( j + 1)( j + 3) ( j + 1)( j + 1)( j + 3)
10( j + 1)( j + 1)( j + 3) 1 1 +
2
b) = = tan
1

( 2 + 1) 2 ( 2 + 9)
2
j + 1 j + 1 j + 3 1+ 2
= R
1+ 2 1+ 2 9 + 2
3 = tan 1 9 +
2
= R (e j1 e j2 e j3 )
3
9 + 2
= 1 + 2 + 3
24
10
j ( j 2 + 2 2 )
10 j (2 2 j 2 )
=
( j 2 + 2 2 ) (2 2 j 2 )
10(2 (2 2 ) j )
c) =
(4 2 + (2 2 ) 2 )
2 (2 2 ) j
=R
4 2 + (2 2 ) 2
j
= R (e )
10 4 2 + (2 2 ) 2 10
R= = ;
(4 + (2 ) ) 4 + (2 2 ) 2
2 2 2 2
2 2
4 2 + (2 2 ) 2
= tan 1 2
4 + (2 2 ) 2
2
1
R= ;
10 2 + 2 1 + 2
2
e 2 j

10 j ( j + 1)( j + 2)
= tan 1 22 + 2
j ( j + 1)( j + 2) 2 j 1 2
= e
d) 10 ( 2 + 1)( 2 + 2) 2 + 2
2
j + 2 j + 1 2 j j / 2
=R e
2 + 1+
2 2 2
2 = tan 1 1 +
2
1
= R (e j1 e j2 e j3 )
1+ 2
= 1 + 2 + 3
MATLABcode:
clear all;
s = tf('s')
Ga=10*(s+2)/(s^2*(s+1)*(s+10))
figure(1)
25
Nyquist(Ga)
Gb=10*s*(s+1)/((s+2)*(s^2+3*s+2))
figure(2)
Nyquist(Gb)
Gc=10*(s+2)/(s*(s^2+2*s+2))
figure(3)
Nyquist(Gc)
Gd=pade(exp(-2*s),1)/(10*s*(s+1)*(s+2))
figure(4)
Nyquist(Gd)
Nyquistplots(polarplots):
Part(a)
26
Nyquist Diagram
15
10
5
Imaginary Axis
-5
-10
-15
-300 -250 -200 -150 -100 -50 0
Real Axis

Part(b)
Nyquist Diagram
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-1 -0.5 0 0.5 1 1.5 2 2.5
Real Axis

Part(c)
27
Nyquist Diagram
80
60
40
20
Imaginary Axis
-20
-40
-60
-80
-7 -6 -5 -4 -3 -2 -1 0
Real Axis

Part(d)
Nyquist Diagram
2.5
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-2.5
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Real Axis

28
2-5) In all cases find the real and imaginary axis intersections.
10 10( j + 2) 10 2 j
G ( j ) = = = ;
( j 2) ( + 4)
2
( 2 + 4) ( 2 + 4)
2
Re {G ( j )} = cos = ,
( 2 + 4)

a) Im {G ( j )} = sin = ,
( 2 + 4)
2
( 2 + 4)
= tan 1
( 2 + 4)
10
R=
( 2 + 4)
lim 0 G ( j ) = 5; = tan 1 1 = 90o

0
lim G ( j ) = 0; = tan 1 0 = 180o
1
Real axis intersection @ j = 0
Imaginary axis int er sec tion does not exist.
b&c) =1 0o
=0 -180o
Therefore:
Re{ G(j) } =
Im {G(j)} =
29
If Re{G(j )} = 0
0
If Im{ G(j )} = 0 0

If = n
90
If = n and = 1
If = n and 0
If = n and 0
d) G(j) =
lim G j = - 90o
lim G j = -180o
e) | |

G(j) = + = tan-1 ( T) L
210
26
MATLABcode:
clear all;
s = tf('s')
%Part(a)
Ga=10/(s-2)
figure(1)
nyquist(Ga)
%Part(b)
zeta=0.5; %asuuming a value for zeta <1
wn=2*pi*10 %asuuming a value for wn
Gb=1/(1+2*zeta*s/wn+s^2/wn^2)
figure(2)
nyquist(Gb)
%Part(c)
zeta=1.5; %asuuming a value for zeta >1
wn=2*pi*10
Gc=1/(1+2*zeta*s/wn+s^2/wn^2)
figure(3)
nyquist(Gc)
%Part(d)
T=3.5 %assuming value for parameter T
Gd=1/(s*(s*T+1))
figure(4)
nyquist(Gd)
211
%Part(e)
T=3.5
L=0.5
Ge=pade(exp(-1*s*L),2)/(s*T+1)
figure(5)
hold on;
nyquist(Ge)
notes:InordertouseMatlabNyquistcommand,parametersneedstobeassignedwithvalues,andPade
approximationneedstobeusedforexponentialterminpart(e).
Nyquistdiagramsareasfollows:
212
Part(a)
Nyquist Diagram
2.5
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-2.5
-5 -4 -3 -2 -1 0 1
Real Axis

Part(b)
Nyquist Diagram
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Real Axis

213
Part(c)
Nyquist Diagram
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Real Axis

Part(d)
214
Nyquist Diagram
60
40
20
Imaginary Axis
-20
-40
-60
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
Real Axis

Part(e)
Nyquist Diagram
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
Real Axis

215
2-7) a) G(j) =
. .
Steps for plotting |G|:
(1) For < 0.1, asymptote is
Break point: = 0.5

Slope = -1 or -20 dB/decade
(2) For 0.5 < < 10
Break point: = 10
Slope = -1+1 = 0 dB/decade
(3) For 10 < < 50:
Break point: = 50
(4) For > 50
Steps for plotting G
(1) = -90o
0: 90
(2) =
0
0: 0
.
(3) =
.
: 90
.
0: 90
.
(4) =
.
0
.
b) Lets convert the transfer function to the following form:
G(j) = . G(s) =
.
Steps for plotting |G|:
(1) Asymptote: < 1 |G(j)| 2.5 /

Slope: -1 or -20 dB/decade
| | 2.5
216
(2) n =2 and = 0.1 for second-order pole

break point: = 2
slope: -3 or -60 dB/decade
| | 5
Steps for plotting G(j):
(1) for term 1/s the phase starts at -90o and at = 2 the phase will be -180o
(2) for higher frequencies the phase approaches -270o
c) Convert the transfer function to the following form:
0.01 1
0.01 1
9
for term , slope is -2 (-40 dB/decade) and passes through | | 1
(1) the breakpoint: = 1 and slope is zero

(2) the breakpoint: = 2 and slope is -2 or -40 dB/decade
|G(j)| = 1 = 2 = 0.01 below the asymptote
|G(j)| = 1 = = = 50 above the asymptote
.
Steps for plotting G:
(1) phase starts from -180o due to
(2) G(j)| =1 = 0
(3) G(j)| = 2 = -180o
d) G(j) =

Steps for plotting the |G|:
(1) Asymptote for <1 is zero
(2) Breakpoint: = 1, slope = -1 or -10 dB/decade
(3) As is a damping ratio, then the magnitude must be obtained for various when
01
217
The high frequency slope is twice that of the asymptote for the single-pole case
Steps for plotting G:
(1) The phase starts at 0o and falls -1 or -20 dB/decade at = 0.2 and approaches -180o
at = 5. For > 5, the phase remains at -180o.
(2) As is a damping ratio, the phase angles must be obtained for various when
01
28)Usethisparttoconfirmtheresultsfromthepreviouspart.
MATLABcode:
s = tf('s')
Ga=2000*(s+0.5)/(s*(s+10)*(s+50))
figure(1)
bode(Ga)
grid on;
Gb=25/(s*(s+2.5*s^2+10))
figure(2)
bode(Gb)
grid on;
Gc=(s+100*s^2+100)/(s^2*(s+25*s^2+100))
figure(3)
bode(Gc)
grid on;
218
zeta = 0.2
wn=8
Gd=1/(1+2*zeta*s/wn+(s/wn)^2)
figure(4)
bode(Gd)
grid on;
t=0.3
'from pade approzimation:'
exp_term=pade(exp(-s*t),1)
Ge=0.03*(exp_term+1)^2/((exp_term-1)*(3*exp_term+1)*(exp_term+0.5))
figure(5)
bode(Ge)
grid on;
Part(a)
219
Bode Diagram
60
40
Magnitude (dB)
20
-20
-40
-60
0
-45
Phase (deg)
-90
-135
-180
-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)

Part(b)
Bode Diagram
50
Magnitude (dB)
-50
-100
-90
-135
Phase (deg)
-180
-225
-270
-1 0 1 2
10 10 10 10
Frequency (rad/sec)

220
Part(c)
Bode Diagram
60
40
Magnitude (dB)
20
-20
-40
0
-45
Phase (deg)
-90
-135
-180
-1 0 1
10 10 10
Frequency (rad/sec)

Part(d)
221
Bode Diagram
10
Magnitude (dB)
-10
-20
-30
-40
-50
0
-45
Phase (deg)
-90
-135
-180
-1 0 1 2
10 10 10 10
Frequency (rad/sec)

Part(e)
Bode Diagram
0
-20
Magnitude (dB)
-40
-60
-80
-100
-120
0
-90
Phase (deg)
-180
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
222
2-9)
a)
1 2 0 0 0
u1 (t )
A = 0 2 3
B = 1 0
u (t ) =
u 2 (t )
1 3 1 0 1
b)
1 2 0 2 0
2 0 1 0 1
3 4 1 0 0
2-10) We know that:

1
2
2

Partial integration of equation (1) gives:

1
sG s g 0 g t
sG s g 0
Differentiation of both sides of equation (1) with respect to s gives:

Comparing with equation (1), we conclude that:
223
2-11) Let g(t) =

then
Using Laplace transform and differentiation property, we have X(s) = sG(s)
Therefore G(s) = , which means:
2-12) By Laplace transform definition:

Now, consider = t - T, then:

Which means:
2-13) Consider:
f(t) = g1(t) g2(t) =
By Laplace transform definition:
By using time shifting theorem, we have:
224
Lets consider g(t) = g1(t) g2(t)
By inverse Laplace Transform definition, we have
1
2
Then
Where
therefore:
1
G s G s
2
2-14) a) We know that
= = sG(s) + g(0)
When s , it can be written as:
225
As
Therefore: 0
b) By Laplace transform differentiation property:
As
Therefore
0 0
which means:
215)
MATLABcode:
clear all;
syms t
s=tf('s')
f1 = (sin(2*t))^2
L1=laplace(f1)
226
% f2 = (cos(2*t))^2 = 1-(sin(2*t))^2 ===> L(f2)=1/s-L(f1) ===>
L2= 1/s - 8/s/(s^2+16)
f3 = (cos(2*t))^2
L3=laplace(f3)
'verified as L2 equals L3'
2
MATLABsolutionfor L sin 2t is: { }
8/s/(s^2+16)
2
{ 2
}
Calculating L cos 2t basedon L sin 2t { }
{ }
L cos 2 2t =(s^^3+8s)/(s^4+16s^2)
2
{
verifying L cos 2t : }
(8+s^2)/s/(s^2+16)
216)(a) (b) (c)
5 4s 1 4
G ( s ) = G ( s ) = + G ( s ) =
( s + 5) 2
(s 2
+4 ) s+2 2
s + 4s + 8
(d) (e)

1 1
G ( s ) = 2
s +4
G ( s) = e kT ( s + 5 )
=
1 e
T ( s +5 )

k =0
227
217)Note:%section (e) requires assignment of T and a numerical loop calculation
MATLABcode:
clear all;
syms t u
f1 = 5*t*exp(-5*t)
L1=laplace(f1)
f2 = t*sin(2*t)+exp(-2*t)
L2=laplace(f2)
f3 = 2*exp(-2*t)*sin(2*t)
L3=laplace(f3)
f4 = sin(2*t)*cos(2*t)
L4=laplace(f4)
%section (e) requires assignment of T and a numerical loop calculation
(a) g (t ) = 5te 5t u s (t )

Answer:5/(s+5)^2
(b) g ( t ) = ( t sin 2t + e 2t )us ( t )
Answer:4*s/(s^2+4)^2+1/(s+2)
(c) g ( t ) = 2e 2t sin 2t u s (t )
Answer:4/(s^2+4*s+8)
228
(d) g ( t ) = sin 2t cos 2t u s (t )
Answer:2/(s^2+16)

(e) g ( t ) = e 5 kT
( t kT ) where(t)=unitimpulsefunction
k =0
%section (e) requires assignment of T and a numerical loop calculation
218(a)
g (t ) = u s (t ) 2u s (t 1) + 2u s ( t 2) 2u s (t 3) + L
s
1 1 e
G(s) = (1 2e s + 2e 2 s 2e 3s + L) =

s (
s 1+ e
s
)
gT (t ) = u s (t ) 2u s (t 1) + u s (t 2) 0t2
1 1
(1 2e s + e 2 s ) = (1 e s )
2
GT ( s ) =
s s
s
1 1 e
g ( t ) = g T
( t 2 k )u s ( t 2 k ) G( s) = s (1 e s 2 2 ks
) e =
s(1 + e
s
)

k =0 k =0
(b)
g (t ) = 2tu s (t ) 4(t 0.5)u s (t 0.5) + 4(t 1)u s (t 1) 4(t 1.5)u s (t 1.5) + L
2 ( 0.5s )
2 1 e
G ( s ) = ( 1 2e
0.5 s
+ 2e
s
2e
1.5 s
) s 2 1 + e0.5s
+L =
s
2
( )
gT ( t ) = 2tu s ( t ) 4( t 0.5)u s ( t 0.5) + 2( t 1)u s ( t 1) 0 t 1
2 2
(1 2e0.5 s + e s ) = s 2 (1 e0.5 s )
2
GT ( s ) = 2

s

2 ( 0.5 s
2 1 e )
g (t ) = gT (t k )us (t k ) G ( s) = (1 e )
0.5 s 2
e
ks
= 2
s (1 + e
k =0 k =0 s
2 0.5 s
)
219)
g ( t ) = ( t + 1)u s ( t ) ( t 1)u s ( t 1) 2u s ( t 1) ( t 2 )u s ( t 2) + ( t 3)u s ( t 3) + u s ( t 3)
229
1 1
G ( s ) =
s
2 (1 e s e 2 s + e 3s ) + s (1 2e s + e 3s )
2-20)
1 1
1
2-21)
1
1
2-22) 0 0 0
0 0
2 2 2 2
230
2 2 2 2
223
MATLABcode:
clear all;
syms t u s x1 x2 Fs
f1 = exp(-2*t)
L1=laplace(f1)/(s^2+5*s+4);
Eq2=solve('s*x1=1+x2','s*x2=-2*x1-3*x2+1','x1','x2')
f2_x1=Eq2.x1
f2_x2=Eq2.x2
f3=solve('(s^3-s+2*s^2+s+2)*Fs=-1+2-(1/(1+s))','Fs')
HereisthesolutionprovidedbyMATLAB:
Part(a):F(s)=1/(s+2)/(s^2+5*s+4)
Part(b):X1(s)=(4+s)/(2+3*s+s^2)
X2(s)=(s2)/(2+3*s+s^2)
Part(c):F(s)=s/(1+s)/(s^3+2*s^2+2)
231
224)
MATLABcode:
clear all;
syms s Fs
f3=solve('s^2*Fs-Fs=1/(s-1)','Fs')
Answer from MATLAB: Y(s)=1/(s-1)/(s^2-1)
225)
MATLABcode:
clear all;
syms s CA1 CA2 CA3
v1=1000;
v2=1500;
v3=100;
k1=0.1
k2=0.2
k3=0.4
f1='s*CA1=1/v1*(1000+100*CA2-1100*CA1-k1*v1*CA1)'
f2='s*CA2=1/v2*(1100*CA1-1100*CA2-k2*v2*CA2)'
f3='s*CA3=1/v3*(1000*CA2-1000*CA3-k3*v3*CA3)'
Sol=solve(f1,f2,f3,'CA1','CA2','CA3')
CA1=Sol.CA1
CA3=Sol.CA2
CA4=Sol.CA3
232
SolutionfromMATLAB:
CA1(s)=
1000*(s*v2+1100+k2*v2)/(1100000+s^2*v1*v2+1100*s*v1+s*v1*k2*v2+1100*s*v2+1100*k2*v2+k1*v1*s*v2+
1100*k1*v1+k1*v1*k2*v2)
CA3(s)=
1100000/(1100000+s^2*v1*v2+1100*s*v1+s*v1*k2*v2+1100*s*v2+1100*k2*v2+k1*v1*s*v2+1100*k1*v1+k1*
v1*k2*v2)
CA4(s)=
1100000000/(1100000000+1100000*s*v3+1000*s*v1*k2*v2+1100000*s*v1+1000*k1*v1*s*v2+1000*k1*v1*k
2*v2+1100*s*v1*k3*v3+1100*s*v2*k3*v3+1100*k2*v2*s*v3+1100*k2*v2*k3*v3+1100*k1*v1*s*v3+1100*k1
*v1*k3*v3+1100000*k1*v1+1000*s^2*v1*v2+1100000*s*v2+1100000*k2*v2+1100000*k3*v3+s^3*v1*v2*v3+
1100*s^2*v1*v3+1100*s^2*v2*v3+s^2*v1*v2*k3*v3+s^2*v1*k2*v2*v3+s*v1*k2*v2*k3*v3+k1*v1*s^2*v2*v3
+k1*v1*s*v2*k3*v3+k1*v1*k2*v2*s*v3+k1*v1*k2*v2*k3*v3)
2-26) (a)
1 1 1 1 1 2t 1 3t
G( s) = + g (t ) = e + e t 0
3s 2( s + 2) 3( s + 3) 3 2 3
(b)
2.5 5 2.5 t t 3t
G( s) = + 2
+ g ( t ) = 2.5e + 5te + 2.5e t 0
s + 1 ( s + 1) s+3
(c)
G(s) = ( 50
s

s +1
20

30 s + 20
s +4
2 ) e
s
[
g ( t ) = 50 20 e
( t 1)
30 cos 2(t 1) 5 sin 2(t 1) ] us (t 1)
(d)
1 s 1 1 1 s
G( s) = 2
= + 2
2
TakingtheinverseLaplacetransform,
s s + s+2 s s + s+2 s + s+2
g (t ) = 1 + 1.069e
0.5t
[sin 1.323t + sin (1.323t 69.3o )] = 1 + e0.5t (1.447 sin 1.323t cos1.323t ) t 0
2 t
(e) g ( t ) = 0.5t e t 0
(f)Try using MATLAB
>> b=num*2
233
b=
2 2 2
>>num =
1 1 1
>> denom1=[1 1]
denom1 =
1 1
>> denom2=[1 5 5]
denom2 =
1 5 5
>> num*2
ans =
2 2 2
>> denom=conv([1 0],conv(denom1,denom2))
denom =
1 6 10 5 0
>> b=num*2
b=
2 2 2
>> a=denom
a=
1 6 10 5 0
>> [r, p, k] = residue(b,a)
r=
-0.9889
234
2.5889
-2.0000
0.4000
p=
-3.6180
-1.3820
-1.0000
k=[]
If there are no multiple roots, then
The number of poles n is
b r r r
= 1 + 2 + ... + n + k
a s + p1 s + p2 s + pn
In this case, p1 and k are zero. Hence,
0.4 0.9889 2.5889 2

G ( s) = +
s s + 3.6180 s + 1.3820 s + 1
g (t ) = 0.4 0.9889e 3.618t + 1.3820e 2.5889t 2e t
(g)
G s 2e 2e 2e 1
(h)
235
(i)
227
MATLABcode:
clear all;
syms s
f1=1/(s*(s+2)*(s+3))
F1=ilaplace(f1)
f2=10/((s+1)^2*(s+3))
F2=ilaplace(f2)
f3=10*(s+2)/(s*(s^2+4)*(s+1))*exp(-s)
F3=ilaplace(f3)
f4=2*(s+1)/(s*(s^2+s+2))
F4=ilaplace(f4)
f5=1/(s+1)^3
F5=ilaplace(f5)
f6=2*(s^2+s+1)/(s*(s+1.5)*(s^2+5*s+5))
F6=ilaplace(f6)
s=tf('s')
f7=(2+2*s*pade(exp(-1*s),1)+4*pade(exp(-2*s),1))/(s^2+3*s+2) %using Pade command

for exponential term
236
[num,den]=tfdata(f7,'v') %extracting the polynomial values
syms s
f7n=(-2*s^3+6*s+12)/(s^4+6*s^3+13*s^2+12*s+4) %generating sumbolic function for

ilaplace
F7=ilaplace(f7n)
f8=(2*s+1)/(s^3+6*s^2+11*s+6)
F8=ilaplace(f8)
f9=(3*s^3+10^s^2+8*s+5)/(s^4+5*s^3+7*s^2+5*s+6)
F9=ilaplace(f9)
SolutionfromMATLABfortheInverseLaplacetransforms:
1
Part(a): G ( s) =
s( s + 2)( s + 3)
G(t)=1/2*exp(2*t)+1/3*exp(3*t)+1/6
Tosimplify:
symst
digits(3)
vpa(1/2*exp(2*t)+1/3*exp(3*t)+1/6)
ans=.500*exp(2.*t)+.333*exp(3.*t)+.167
10
Part(b): G ( s) =
( s + 1) 2 ( s + 3)
G(t)=5/2*exp(3*t)+5/2*exp(t)*(1+2*t)
237
100( s + 2)
Part(c): G( s) = 2
e s
s( s + 4 )( s + 1)
G(t)=Step(t1)*(4*cos(t1)^2+2*sin(t1)*cos(t1)+4*exp(1/2*t+1/2)*cosh(1/2*t1/2)4*exp(t+1)cos(2*t2)
2*sin(2*t2)+5)
2( s + 1)
Part(d): G ( s) =
s( s 2 + s + 2 )
G(t)=1+1/7*exp(1/2*t)*(7*cos(1/2*7^(1/2)*t)+3*7^(1/2)*sin(1/2*7^(1/2)*t))
Tosimplify:
symst
digits(3)
vpa(1+1/7*exp(1/2*t)*(7*cos(1/2*7^(1/2)*t)+3*7^(1/2)*sin(1/2*7^(1/2)*t)))
ans=1.+.143*exp(.500*t)*(7.*cos(1.32*t)+7.95*sin(1.32*t))
1
Part(e): G ( s) =
( s + 1) 3
G(t)=1/2*t^2*exp(t)
2( s 2 + s + 1)
Part(f): G( s) =
. )( s 2 + 5s + 5)
s( s + 15
G(t)=4/15+28/3*exp(3/2*t)16/5*exp(5/2*t)*(3*cosh(1/2*t*5^(1/2))+5^(1/2)*sinh(1/2*t*5^(1/2)))
2 + 2 se s + 4e 2 s
Part(g):G ( s ) =
s 2 + 3s + 2
238
G(t)=2*exp(2*t)*(7+8*t)+8*exp(t)*(2+t)
2s + 1
Part(h): G ( s) =
s + 6 s 2 + 11s + 6
3
G(t)=1/2*exp(t)+3*exp(2*t)5/2*exp(3*t)
3s 3 + 10 s 2 + 8s + 5
Part(i): G ( s) = 4
s + 5 s 3 + 7 s 2 + 5s + 6
G(t)= 7*exp(2*t)+10*exp(3*t)
1/10*ilaplace(10^(2*s)/(s^2+1)*s,s,t)+1/10*ilaplace(10^(2*s)/(s^2+1),s,t)+1/10*sin(t)*(10+dirac(t)*(exp(
3*t)+2*exp(2*t)))
239
2-28) Ax t Bu t
a)
b)
3 2
2 2
2-29) (a) (b)

Y (s) 3s + 1 Y (s) 5
= 3 2
= 4 2

R( s) s + 2 s + 5s + 6 R( s) s + 10 s + s + 5
(c) (d)
s
Y (s) s ( s + 2) Y (s) 1 + 2e
= 4 3 2
= 2

R( s) s + 10 s + 2 s + s + 2 R( s) 2s + s + 5
e) 1
5 2 2
By using Laplace transform, we have:
4 5 2
As X s e Y s , then
240
2 1
4
Then:
f) By using Laplace transform we have:
2
2 2 2
As a result:
230)
AftertakingtheLaplacetransform,theequationwassolvedintermsofY(s),andconsecutivelywasdividedby
inputR(s)toobtainY(s)/R(s):
MATLABcode:
clear all;
syms Ys Rs s
sol1=solve('s^3*Ys+2*s^2*Ys+5*s*Ys+6*Ys=3*s*Rs+Rs','Ys')
Ys_Rs1=sol1/Rs
sol2=solve('s^4*Ys+10*s^2*Ys+s*Ys+5*Ys=5*Rs','Ys')
Ys_Rs2=sol2/Rs
sol3=solve('s^3*Ys+10*s^2*Ys+2*s*Ys+2*Ys/s=s*Rs+2*Rs','Ys')
Ys_Rs3=sol3/Rs
sol4=solve('2*s^2*Ys+s*Ys+5*Ys=2*Rs*exp(-1*s)','Ys')
Ys_Rs4=sol4/Rs
241
%Note: Parts E&F are too complicated with MATLAB, Laplace of integral is not
executable in MATLAB.....skipped
MATLABAnswers:
Part(a): Y(s)/R(s)=(3*s+1)/(5*s+6+s^3+2*s^2);
Part(b): Y(s)/R(s)=5/(10*s^2+s+5+s^4)
Part(c): Y(s)/R(s)=(s+2)*s/(2*s^2+2+s^4+10*s^3)
Part(d): Y(s)/R(s)=2*exp(s)/(2*s^2+s+5)
%Note: Parts E&F are too complicated with MATLAB, Laplace of integral is not
executable in MATLAB.....skipped
231
MATLABcode:
clear all;
s=tf('s')
%Part a
Eq=10*(s+1)/(s^2*(s+4)*(s+6));
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part b
Eq=(s+1)/(s*(s+2)*(s^2+2*s+2));
%Part c
Eq=5*(s+2)/(s^2*(s+1)*(s+5));
242
%Part d
Eq=5*(pade(exp(-2*s),1))/(s^2+s+1); %Pade approximation oreder 1 used
%Part e
Eq=100*(s^2+s+3)/(s*(s^2+5*s+3));
%Part f
Eq=1/(s*(s^2+1)*(s+0.5)^2);
%Part g
Eq=(2*s^3+s^2+8*s+6)/((s^2+4)*(s^2+2*s+2));
%Part h
Eq=(2*s^4+9*s^3+15*s^2+s+2)/(s^2*(s+2)*(s+1)^2);
Thesolutionsarepresentedintheformoftwovectors,randp,whereforeachcase,thepartialfraction
expansionisequalto:
243
b( s ) r r r
= 1 + 2 + ... + n
a( s ) s p1 s p 2 s pn
Followingarerandpvectorsforeachpart:
Part(a):
r=0.6944
0.9375
0.2431
0.4167
p=6.0000
4.0000
Part(b):
r=0.2500
0.25000.0000i
0.2500+0.0000i
0.2500
p=2.0000
1.0000+1.0000i
244
1.00001.0000i
Part(c):
r=0.1500
1.2500
1.4000
2.0000
p=5
Part(d):
r=10.0000
5.00000.0000i
5.0000+0.0000i
p=1.0000
0.5000+0.8660i
0.50000.8660i
245
Part(e):
r=110.9400
110.9400
100.0000
p=4.3028
0.6972
Part(f):
r=0.2400+0.3200i
0.24000.3200i
4.4800
1.6000
4.0000
p=0.0000+1.0000i
0.00001.0000i
0.5000
0.5000
246
Part(g):
r=0.1000+0.0500i
0.10000.0500i
1.1000+0.3000i
1.10000.3000i
p=0.0000+2.0000i
0.00002.0000i
1.0000+1.0000i
1.00001.0000i
Part(h):
r=5.0000
1.0000
9.0000
2.0000
1.0000
p=2.0000
1.0000
1.0000
247
232)
MATLABcode:
clear all;
syms s
%Part a
Eq=10*(s+1)/(s^2*(s+4)*(s+6));
ilaplace(Eq)
%Part b
Eq=(s+1)/(s*(s+2)*(s^2+2*s+2));
ilaplace(Eq)
%Part c
Eq=5*(s+2)/(s^2*(s+1)*(s+5));
ilaplace(Eq)
%Part d
exp_term=(-s+1)/(s+1) %pade approcimation
Eq=5*exp_term/((s+1)*(s^2+s+1));
ilaplace(Eq)
%Part e
Eq=100*(s^2+s+3)/(s*(s^2+5*s+3));
ilaplace(Eq)
%Part f
Eq=1/(s*(s^2+1)*(s+0.5)^2);
ilaplace(Eq)
248
%Part g
Eq=(2*s^3+s^2+8*s+6)/((s^2+4)*(s^2+2*s+2));
ilaplace(Eq)
%Part h
Eq=(2*s^4+9*s^3+15*s^2+s+2)/(s^2*(s+2)*(s+1)^2);
ilaplace(Eq)
MATLABAnswers:
Part(a):
G(t)=15/16*exp(4*t)+25/36*exp(6*t)+35/144+5/12*t
Tosimplify:
symst
digits(3)
vpa(15/16*exp(4*t)+25/36*exp(6*t)+35/144+5/12*t)
ans=.938*exp(4.*t)+.694*exp(6.*t)+.243+.417*tPart(b):
G(t)=1/4*exp(2*t)+1/41/2*exp(t)*cos(t)
Part(c):
G(t)=5/4*exp(t)7/5+3/20*exp(5*t)+2*t
249
Part(d):
G(t)=5*exp(1/2*t)*(cos(1/2*3^(1/2)*t)+3^(1/2)*sin(1/2*3^(1/2)*t))+5*(1+2*t)*exp(t)
Part(e):
G(t)=100800/13*exp(5/2*t)*13^(1/2)*sinh(1/2*t*13^(1/2))
Part(f):
G(t)=4+12/25*cos(t)16/25*sin(t)8/25*exp(1/2*t)*(5*t+14)
Part(g):
G(t)=1/5*cos(2*t)1/10*sin(2*t)+1/5*(11*cos(t)3*sin(t))*exp(t)
Part(h):
G(t)=2+t+5*exp(2*t)+(1+9*t)*exp(t)
250
2-33)(a)Polesareat s = 0, 15
. + j16583
. , 15
. j16583
. Onepolesats=0.Marginallystable.
(b)Polesareat s = 5, j 2 , j 2 Twopoleson j axis.Marginallystable.
(c)Polesareat s = 0.8688, 0.4344 + j 2.3593, 0.4344 j 2.3593TwopolesinRHP.Unstable.
(d)Polesareat s = 5, 1 + j , 1 j AllpolesintheLHP.Stable.
(e)Polesareat s = 13387
. , 16634
. + j 2.164, 16634
. j 2.164 TwopolesinRHP.Unstable.
(f)Polesareat s = 22.8487 j 22.6376, 213487

. j 22.6023 TwopolesinRHP.Unstable.
2-34) Find the Characteristic equations and then use the roots command.
(a)
p= [ 1 3 5 0]
sr = roots(p)
p=
1 3 5 0
sr =
-1.5000 + 1.6583i
-1.5000 - 1.6583i
(b) p=conv([1 5],[1 0 2])
sr = roots(p)
p=
1 5 2 10
251
sr =
-5.0000
0.0000 + 1.4142i
0.0000 - 1.4142i
(c)
>> roots([1 5 5])
ans =
-3.6180
-1.3820
(d) roots(conv([1 5],[1 2 2]))
ans =
-5.0000
-1.0000 + 1.0000i
-1.0000 - 1.0000i
(e) roots([1 -2 3 10])
ans =
1.6694 + 2.1640i
1.6694 - 2.1640i
-1.3387
252
(f) roots([1 3 50 1 10^6])
-22.8487 +22.6376i
-22.8487 -22.6376i
21.3487 +22.6023i
21.3487 -22.6023i
Alternatively
Problem234
MATLABcode:
% Question 2-34,
clear all;
s=tf('s')
%Part a
Eq=10*(s+2)/(s^3+3*s^2+5*s);
roots(den)
%Part b
Eq=(s-1)/((s+5)*(s^2+2));
roots(den)
%Part c
Eq=1/(s^3+5*s+5);
roots(den)
253
%Part d
Eq=100*(s-1)/((s+5)*(s^2+2*s+2));
roots(den)
%Part e
Eq=100/(s^3-2*s^2+3*s+10);
roots(den)
%Part f
Eq=10*(s+12.5)/(s^4+3*s^3+50*s^2+s+10^6);
roots(den)
MATLABanswer:
Part(a)

1.5000+1.6583i
1.50001.6583i
Part(b)

5.0000
254
0.0000+1.4142i
0.00001.4142i
Part(c)

0.4344+2.3593i
0.43442.3593i
0.8688
Part(d)

5.0000
1.0000+1.0000i
1.00001.0000i
Part(e)

1.6694+2.1640i
1.66942.1640i
1.3387
Part(f)

255
22.8487+22.6376i
22.848722.6376i
21.3487+22.6023i
21.348722.6023i
2-35)
(a) s 3 + 25 s 2 + 10 s + 450 = 0 Roots: 25.31, 0.1537 + j 4.214, 0.1537 4.214
RouthTabulation:
3
s 1 10
2
s 25 450
250 450 Twosignchangesinthefirstcolumn.TworootsinRHP.
1
s = 8 0
25
0
s 450
3 2
(b) s + 25 s + 10 s + 50 = 0 Roots: 24.6769, 0.1616 + j1.4142, 0.1616 j1.4142
RouthTabulation:
3
s 1 10
2
s 25 50
250 50 Nosignchangesinthefirstcolumn.NorootsinRHP.
1
s =8 0
25
0
s 50
(c) s 3 + 25 s 2 + 250 s + 10 = 0 Roots: 0.0402, 12.48 + j 9.6566, j 9.6566
RouthTabulation:
256
3
s 1 250
2
s 25 10
6250 10 Nosignchangesinthefirstcolumn.NorootsinRHP.
1
s = 249.6 0
25
0
s 10
4 3 2
(d) 2 s + 10 s + 5.5 s + 5.5 s + 10 = 0 Roots: 4.466, 1116
. , 0.2888 + j 0.9611, 0.2888 j 0.9611
RouthTabulation:
4
s 2 5.5 10
3
s 10 5.5
2 55 11
s = 4.4 10
10
1 24.2 100
s = 75.8
4.4
0
s 10
Twosignchangesinthefirstcolumn.TworootsinRHP.
(e) s 6 + 2 s 5 + 8 s 4 + 15 s 3 + 20 s 2 + 16 s + 16 = 0 Roots: 1.222 j 0.8169, 0.0447 j1153

. , 0.1776 j 2.352
RouthTabulation:
6
s 1 8 20 16
5
s 2 15 16
16 15 40 16
4
s = 0.5 = 12
2 2
3
s 33 48
257
2 396 + 24
s = 11.27 16
33
1 5411
. + 528
s = 116
. 0
11.27
0
s 0
Foursignchangesinthefirstcolumn.FourrootsinRHP.
4 3 2
(f) s + 2 s + 10 s + 20 s + 5 = 0 Roots: 0.29, 1788
. , 0.039 + j 3105
. , 0.039 j 3105
.
RouthTabulation:
4
s 1 10 5
3
s 2 20
20 20
2
s =0 5
2
2
s 5 Replace 0 in last row by
1 20 10 10
s
Twosignchangesinfirstcolumn.TworootsinRHP.
0
s 5
(g)
s8 1 8 20 16 0
s7 2 12 16 0 0
s6 2 12 16 0 0
s5 0 0 0 0 0
2 12 16
12 60 64
258
s5 12 60 64 0
s4 2 16 0 0
3
s3 28 64 0 0
s2 0.759 0 0 0
s1 28 0
s0 0
2-36) Use MATLAB roots command
a) roots([1 25 10 450])
ans =
-25.3075
0.1537 + 4.2140i
0.1537 - 4.2140i
b) roots([1 25 10 50])
ans =
-24.6769
-0.1616 + 1.4142i
-0.1616 - 1.4142i
c) roots([1 25 250 10])
ans =
-12.4799 + 9.6566i
-12.4799 - 9.6566i
259
-0.0402
d) roots([2 10 5.5 5.5 10])
ans =
-4.4660
-1.1116
0.2888 + 0.9611i
0.2888 - 0.9611i
e) roots([1 2 8 15 20 16 16])
ans =
0.1776 + 2.3520i
0.1776 - 2.3520i
-1.2224 + 0.8169i
-1.2224 - 0.8169i
0.0447 + 1.1526i
0.0447 - 1.1526i
f) roots([1 2 10 20 5])
ans =
0.0390 + 3.1052i
0.0390 - 3.1052i
-1.7881
-0.2900
g) roots([1 2 8 12 20 16 16])
260
ans =
0.0000 + 2.0000i
0.0000 - 2.0000i
-1.0000 + 1.0000i
-1.0000 - 1.0000i
0.0000 + 1.4142i
0.0000 - 1.4142i
AlternativelyusetheapproachinthisChaptersSection214:
1. ActivateMATLAB
2. GotothedirectorycontainingtheACSYSsoftware.
3. Typein
Acsys

4. ThenpressthetransferfunctionSymbolicbutton
261

5. EnterthecharacteristicequationinthedenominatorandpresstheRouthHurwitzpush
button.
RH =
[ 1, 10]
[ 25, 450]
[ -8, 0]
[ 450, 0]
Two sign changes in the first column. Two roots in RHP=> UNSTABLE
2-37) Use the MATLAB roots command same as in the previous problem.
262
2-38) To solve using MATLAB, set the value of K in an iterative process and find the roots such that at
least one root changes sign from negative to positive. Then increase resolution if desired.
Example: in this case 0<K<12 ( increase resolution by changing the loop to: for K=11:.1:12)
forK=0:12
roots([1251520K])
end
K=
ans =
-24.4193
-0.2904 + 0.8572i
-0.2904 - 0.8572i
K=
ans =
263
-24.4192
-0.2645 + 0.8485i
-0.2645 - 0.8485i
-0.0518
K=
ans =
-24.4191
-0.2369 + 0.8419i
-0.2369 - 0.8419i
-0.1071
K=
264
ans =
-24.4191
-0.2081 + 0.8379i
-0.2081 - 0.8379i
-0.1648
K=
ans =
-24.4190
-0.1787 + 0.8369i
-0.1787 - 0.8369i
-0.2237
K=
265
ans =
-24.4189
-0.1496 + 0.8390i
-0.1496 - 0.8390i
-0.2819
K=
ans =
-24.4188
-0.1215 + 0.8438i
-0.1215 - 0.8438i
-0.3381
K=
266
ans =
-24.4188
-0.0951 + 0.8508i
-0.0951 - 0.8508i
-0.3911
K=
ans =
-24.4187
-0.0704 + 0.8595i
-0.0704 - 0.8595i
-0.4406
K=
267
ans =
-24.4186
-0.0475 + 0.8692i
-0.0475 - 0.8692i
-0.4864
K=
10
ans =
-24.4186
-0.0263 + 0.8796i
-0.0263 - 0.8796i
-0.5288
K=
268
11
ans =
-24.4185
-0.0067 + 0.8905i
-0.0067 - 0.8905i
-0.5681
K=
12
ans =
-24.4184
0.0115 + 0.9015i
0.0115 - 0.9015i
-0.6046
AlternativelyusetheapproachinthisChaptersSection214:
269
1. ActivateMATLAB
3. Typein
Acsys

4. ThenpressthetransferfunctionSymbolicbutton
270

button.
RH=
[1,15,k]
[25,20,0]
[71/5,k,0]
[125/71*k+20,0,0]
[k,0,0]
6. FindthevaluesofKtomakethesystemunstablefollowingthenextsteps.
271
AlternativeProblem236
UsingACSYStoolbarunderTransferFunctionSymbolic,theRouthHurwitzoptioncanbeusedtogenerateRH
matrixbasedondenominatorpolynhomial.Thesystemisstableifandonlyifthefirstcolumnofthismatrix
containsNOnegativevalues.
MATLABcode:tocalculatethenumberofrighthandsidepoles
%Part a
den_a=[1 25 10 450]
roots(den_a)
%Part b
den_b=[1 25 10 50]
roots(den_b)
%Part c
den_c=[1 25 250 10]
roots(den_c)
%Part d
den_d=[2 10 5.5 5.5 10]
roots(den_d)
%Part e
den_e=[1 2 8 15 20 16 16]
roots(den_e)
%Part f
den_f=[1 2 10 20 5]
roots(den_f)
272
%Part g
den_g=[1 2 8 12 20 16 16 0 0]
roots(den_g)
usingACSYS,thedenominatorpolynomialcanbeinserted,andbyclickingontheRouthHurwitzbutton,theR
HchartcanbeobservedinthemainMATLABcommandwindow:
Part(a):forthetransferfunctioninpart(a),thischartis:
RHchart=
[1,10]
[25,450]
[8,0]
[450,0]
Unstablesystemdueto8onthe3rdrow.
2complexconjugatepolesonrighthandside.Allthepolesare:
25.3075
0.1537+4.2140iand0.15374.2140i
273
Part(b):
RHchart:
[1,10]
[25,50]
[8,0]
[50,0]
Stablesystem>>Norighthandsidepole
Part(c):
274
RHchart:
[1,250]
[25,10]
[1248/5,0]
[10,0]
Part(d):
RHchart:
[2,11/2,10]
[10,11/2,0]
[22/5,10,0]
[379/22,0,0]
[10,0,0]
Unstablesystemdueto379/22onthe4throw.
4.4660
1.1116
0.2888+0.9611i
0.28880.9611i
275
Part(e):
RHchart:
[1,8,20,16]
[2,15,16,0]
[1/2,12,16,0]
[33,48,0,0]
[124/11,16,0,0]
[36/31,0,0,0]
[16,0,0,0]
Unstablesystemdueto33and36/31onthe4thand6throw.
0.1776+2.3520i
0.17762.3520i
1.2224+0.8169i
1.22240.8169i
0.0447+1.1526i
0.04471.1526i
Part(f):
RHchart:
[1,10,5]
[2,20,0]
276
[eps,5,0]
[(10+20*eps)/eps,0,0]
[5,0,0]
Unstablesystemdueto((10+20*eps)/eps)onthe4th.
2complexconjugatepolesslightlyonrighthandside.Allthepolesare:
0.0390+3.1052i
0.03903.1052i
1.7881
0.2900
Part(g):
RHchart:
[1,8,20,16,0]
[2,12,16,0,0]
[2,12,16,0,0]
[12,48,32,0,0]
[4,32/3,0,0,0]
[16,32,0,0,0]
[8/3,0,0,0,0]
[32,0,0,0,0]
[0,0,0,0,0]
277
6poleswtzerorealpart:
0.0000+2.0000i
0.00002.0000i
1.0000+1.0000i
1.00001.0000i
0.0000+1.4142i
0.00001.4142i
278
(a) s 4 + 25 s 3 + 15 s 2 + 20 s + K = 0
RouthTabulation:
4
s 1 15 K
3
s 25 20
2 375 20
s = 14.2 K
25
1 284 25 K
s = 20 176
. K 20 176
. K > 0 or K < 1136
.
14.2
0
s K K>0
Thus,thesystemisstablefor0<K<11.36.WhenK=11.36,thesystemismarginallystable.The
2 2
auxiliaryequationis A( s ) = 14.2 s + 11.36 = 0. ThesolutionofA(s)=0is s = 0.8. The
frequencyofoscillationis0.894rad/sec.
4 3 2
(b) s + Ks + 2 s + ( K + 1) s + 10 = 0
RouthTabulation:
4
s 1 2 10
3
s K K +1 K>0
2 2K K 1 K 1
s = 10 K >1
K K
2
1 9 K 1 2
s 9K 1 > 0
K 1
0
s 10
2 2
Theconditionsforstabilityare:K>0,K>1,and 9 K 1 > 0 .Since K isalwayspositive,the
lastconditioncannotbemetbyanyrealvalueofK.Thus,thesystemisunstableforallvaluesofK.
279
3 2
(c) s + ( K + 2 ) s + 2 Ks + 10 = 0
RouthTabulation:
3
s 1 2K
2
s K+2 10 K > 2
2
2 K + 4 K 10
1 2
s K + 2K 5 > 0
K+2
0
s 10
2
Theconditionsforstabilityare:K>2and K + 2 K 5 > 0 or(K+3.4495)(K1.4495)>0,
orK>1.4495.Thus,theconditionforstabilityisK>1.4495.WhenK=1.4495thesystemis
2 2
marginallystable.Theauxiliaryequationis A( s ) = 3.4495 s + 10 = 0. Thesolutionis s = 2.899 .
Thefrequencyofoscillationis1.7026rad/sec.
3 2
(d) s + 20 s + 5 s + 10 K = 0
RouthTabulation:
3
s 1 5
2
s 20 10 K
100 10 K
1
s = 5 0.5K 5 0.5K > 0 or K < 10
20
0
s 10 K K>0
Theconditionsforstabilityare:K>0andK<10.Thus,0<K<10.WhenK=10,thesystemis
2
marginallystable.Theauxiliaryequationis A( s ) = 20 s + 100 = 0.Thesolutionoftheauxiliary
2
equationis s = 5. Thefrequencyofoscillationis2.236rad/sec.
280
4 3 2
(e) s + Ks + 5 s + 10 s + 10 K = 0
RouthTabulation:
4
s 1 5 10 K
3
s K 10 K>0
2 5 K 10
s 10 K 5K 10 > 0 or K > 2
K
50 K 100 2
10 K 3
1 K 50 K 100 10 K 3
s = 5 K 10 K > 0
5K 10 5 K 10
K
0
s 10 K K>0
3
Theconditionsforstabilityare:K>0,K>2,and 5 K 10 K > 0.
UseMatlabtosolveforkfromlastcondition
>>symsk
>>kval=solve(5*k10+k^3,k);
>>eval(kval)
kval=
1.4233
0.7117+2.5533i
0.71172.5533i
SoK>1.4233.
Thus,theconditionsforstabilityis:K>2
281
4 3 2
(f) s + 12.5 s + s + 5 s + K = 0
RouthTabulation:
4
s 1 1 K
3
s 12.5 5
2 12.5 5
s = 0.6 K
12.5
1 3 12.5 K
s = 5 20.83 K 5 20.83 K > 0 or K < 0.24
0.6
0
s K K>0
Theconditionforstabilityis0<K<0.24.WhenK=0.24thesystemismarginallystable.Theauxiliary
2 2
equationis A( s ) = 0.6 s + 0.24 = 0. Thesolutionoftheauxiliaryequationis s = 0.4. Thefrequencyof
oscillationis0.632rad/sec.
2-39)
3 2
Thecharacteristicequationis Ts + ( 2T + 1) s + ( 2 + K ) s + 5 K = 0
RouthTabulation:
3
s T K+2 T >0

2
s 2T + 1 5K T > 1 / 2
1 ( 2T + 1)( K + 2 ) 5 KT
s K (1 3T ) + 4T + 2 > 0
2T + 1
0
s 5K K>0
4T + 2
Theconditionsforstabilityare: T>0,K>0,and K < .Theregionsofstabilityinthe
3T 1
282
TversusKparameterplaneisshownbelow.
240UsetheapproachinthisChaptersSection214:
1. ActivateMATLAB
3. Typein
Acsys
283

4. ThenpressthetransferfunctionSymbolicbutton.
284
button.
RH=
[1,50000,24*k]
[600,k,80*k]
[1/600*k+50000,358/15*k,0]
[(35680*k1/600*k^2)/(1/600*k+50000),80*k,0]
[24*k*(k^221622400*k+5000000000000)/(k30000000)/(35680*k1/600*k^2)*(1/600*k+50000),
0,0]
[80*k,0,0]
6. FindthevaluesofKtomakethesystemunstablefollowingthenextsteps.
(a)Characteristicequation: s 5 + 600 s 4 + 50000 s 3 + Ks 2 + 24 Ks + 80 K = 0
RouthTabulation:
5
s 1 50000 24 K
4
s 600 K 80 K
7
3 3 10 K 14320 K 7
s K < 3 10
600 600
2
2 21408000 K K
s 7
80 K K < 21408000
3 10 K
16 11 2
1 7.2 10 + 3113256
. 10 K 14400 K 2 7 12
s K 2.162 10 K + 5 10 < 0
600(21408000 K )
0
s 80 K K>0
285
Conditionsforstability:
3 7
Fromthe s row: K < 3 10
2 7
Fromthe s row: K < 2.1408 10
1 2 7 12 5 7
Fromthe s row: K 2.162 10 K + 5 10 < 0 or ( K 2.34 10 )( K 2.1386 10 ) < 0
5 7
Thus, 2.34 10 < K < 2.1386 10
0
Fromthe s row: K>0
5 7
Thus,thefinalconditionforstabilityis: 2.34 10 < K < 2.1386 10
5
When K = 2.34 10 = 10.6 rad/sec.
7
When K = 2.1386 10 = 188.59 rad/sec.
(b)Characteristicequation: s 3 + ( K + 2 ) s 2 + 30 Ks + 200 K = 0
Routhtabulation:
3
s 1 30 K
2
s K+2 200 K K > 2
2
1 30 K 140 K
s K > 4.6667
K+2
0
s 200 K K>0
StabilityCondition:K>4.6667
2 2
WhenK=4.6667,theauxiliaryequationis A( s ) = 6.6667 s + 933.333 = 0 .Thesolutionis s = 140.
3 2
(c)Characteristicequation: s + 30 s + 200 s + K = 0
286
Routhtabulation:
3
s 1 200
2
s 30 K
6000 K
1
s K < 6000
30
0
s K K>0
StabililtyCondition: 0 < K < 6000
2 2
WhenK=6000,theauxiliaryequationis A( s ) = 30 s + 6000 = 0. Thesolutionis s = 200.
3 2
(d)Characteristicequation: s + 2 s + ( K + 3) s + K + 1 = 0
Routhtabulation:
3
s 1 K+3
2
s 2 K +1
K +5
1
s K > 5
30
0
s K +1 K > 1
Stabilitycondition: K>1.WhenK=1thezeroelementoccursinthefirstelementofthe
0
s row.Thus,thereisnoauxiliaryequation.WhenK=1,thesystemismarginallystable,andone
ofthethreecharacteristicequationrootsisats=0.Thereisnooscillation.Thesystemresponse
wouldincreasemonotonically.
287
242Stateequation: Openloopsystem: x& ( t ) = Ax ( t ) + Bu ( t )
1 2 0
A= B=
10 0 1
Closedloopsystem: x& ( t ) = ( A BK )x ( t )
1 2
A BK =
10 k1 k 2
Characteristicequationoftheclosedloopsystem:
s 1 2
sI A + BK = = s + ( k 2 1) s + 20 2k1 k 2 = 0
2

10 + k1 s + k2
Stabilityrequirements:
k 2 1 > 0 or k 2 > 1
20 2k1 k 2 > 0 or k 2 < 20 2k1
Parameterplane:
243)Characteristicequationofclosedloopsystem:
s 1 0
sI A + BK = 0 1 = s + ( k 3 + 3 ) s + ( k 2 + 4 ) s + k1 = 0
3 2
s
k1 k2 + 4 s + k3 + 3
288
RouthTabulation:
k2 + 4
3
s 1
k3 + 3 k 3 +3>0 or k 3 > 3
2
s k1
(k + 3 )( k 2 + 4 ) k1
s
1 3
(k + 3) ( k + 4 ) k > 0
k3 + 3
3 2 1
k >0
0
s k 1 1
StabilityRequirements:
k3 > 3, k1 > 0, (k 3
+ 3 )( k 2 + 4 ) k1 > 0
244(a)SinceAisadiagonalmatrixwithdistincteigenvalues,thestatesaredecoupledfromeachother.The
secondrowofBiszero;thus,thesecondstatevariable, x 2 isuncontrollable.Sincetheuncontrollable
statehastheeigenvalueat3whichisstable,andtheunstablestate x3 withtheeigenvalueat2is
controllable,thesystemisstabilizable.
(b)Sincetheuncontrollablestate x1hasanunstableeigenvalueat1,thesystemisnostabilizable.
2-45) a)
If , then or
If , then 1 . As a result:
1
1 1
289
1
b
As a result:
1 1
Y ( s) G ( s) H (s) ( K p + K d s)
= =
X ( s ) (1 + G ( s ) H ( s )) (( s + 1)( s 2 g / l ) + K p + K d s )
( K p + K d s)
=
( s + ( ( g / l ) + 1) s 2 + K d s g / l + K p )
3
c) lets choose 10 0.1.
UsetheapproachinthisChaptersSection214:
1. ActivateMATLAB
3. Typein
Acsys
290

291

button.
RH =
[ 1/10, kd]
[ eps, kp-10]
[ (-1/10*kp+1+kd*eps)/eps, 0]
[ kp-10, 0]
For the choice of g/l or the system will be unstable. The quantity g/l must be >1.
Increase g/l to 1.1 and repeat the process.
292
d) Use the ACSYS toolbox as in section 2-14 to find the inverse Laplace transform. Then plot
the time response by selecting the parameter values. Or use toolbox 2-6-1.
UsetheapproachinthisChaptersSection214:
1. ActivateMATLAB
3. Typein
Acsys

293

5. EnterthecharacteristicequationinthedenominatorandpresstheInverseLaplaceTransform
pushbutton.
----------------------------------------------------------------
Inverse Laplace Transform
----------------------------------------------------------------
294
G(s) =
[ kd kp ]
[------------------------ ------------------------------]
[ 3 3 ]
[1/10 s + s kd + kp - 10 1/10 s + s kd + kp - 10]
G(s) factored:
[ kd kp ]
[10 -------------------------- 10 --------------------------]
[ 3 3 ]
[ s + 10 s kd + 10 kp - 100 s + 10 s kd + 10 kp - 100]
Inverse Laplace Transform:
g(t) =
matrix([[10*kd*sum(1/(3*_alpha^2+10*kd)*exp(_alpha*t),_alpha=RootOf(_Z^3+10*_Z*kd
+10*kp-
100)),10*kp*sum(1/(3*_alpha^2+10*kd)*exp(_alpha*t),_alpha=RootOf(_Z^3+10*_Z*kd+1
0*kp-100))]])
While MATLAB is having a hard time with this problem, it is easy to see the solution will be unstable
for all values of Kp and Kd. Stability of a linear system is independent of its initial conditions. For
different values of g/l and , you may solve the problem similarly assign all values (including Kp and
Kd) and then find the inverse Laplace transform of the system. Find the time response and apply the
initial conditions.
Lets chose g/l=1 and keep =0.1, take Kd=1 and Kp=10.
295
Y (s) G (s) H (s) ( K p + K d s)

= =
X ( s ) (1 + G ( s ) H ( s )) (( s + 1)( s 2 g / l ) + K p + K d s )
(10 + s ) (10 + s )
= =
(0.1s + (0.1(1) + 1) s + s 1 + 10) (0.1s + 0.9 s 2 + s + 9)
3 2 3
Using ACSYS:
RH =
[ 1/10, 1]
[ 9/10, 9]
[ 9/5, 0]
[ 9, 0]
Hence the system is stable
----------------------------------------------------------------
Inverse Laplace Transform
----------------------------------------------------------------
G(s) =
s + 10
-------------------------
3 2
1/10 s + 9/10 s + s + 9
296
G factored:
Zero/pole/gain:
10 (s+10)
-----------------
(s+9) (s^2 + 10)
Inverse Laplace Transform:
g(t) = -10989/100000*exp(-
2251801791980457/40564819207303340847894502572032*t)*cos(79057/25000*t)+868757373/25000
0000*exp(-
2251801791980457/40564819207303340847894502572032*t)*sin(79057/25000*t)+10989/100000*ex
p(-9*t)
Use this MATLAB code to plot the time response:
fori=1:1000
t=0.1*i;
tf(i)=10989/100000*exp(
2251801791980457/40564819207303340847894502572032*t)*cos(79057/25000*t)+868757373/250
000000*exp(
2251801791980457/40564819207303340847894502572032*t)*sin(79057/25000*t)+10989/100000*e
xp(9*t);
end
figure(3)
plot(1:1000,tf)
297

252)USEMATLAB
symst
f=5+2*exp(2*t)*sin(2*t+pi/4)4*exp(2*t)*cos(2*tpi/2)+3*exp(4*t)
F=laplace(f)
cltF=F/(1+F)
f=
5+2*exp(2*t)*sin(2*t+1/4*pi)4*exp(2*t)*sin(2*t)+3*exp(4*t)
F=
(8*s^3+44*s^2+112*s+160+8*2^(1/2)*s^2+16*2^(1/2)*s+2^(1/2)*s^3)/s/(s^2+4*s+8)/(s+4)
cltF=
298
(8*s^3+44*s^2+112*s+160+8*2^(1/2)*s^2+16*2^(1/2)*s+2^(1/2)*s^3)/s/(s^2+4*s+8)/(s+4)/(1+(8*s^3
+44*s^2+112*s+160+8*2^(1/2)*s^2+16*2^(1/2)*s+2^(1/2)*s^3)/s/(s^2+4*s+8)/(s+4))
symss
cltFsimp=simplify(cltF)
NexttypethedenominatorintoACSYSRouthHurwitzprogram.
char=collect(s^4+16*s^3+68*s^2+144*s+160+8*2^(1/2)*s^2+16*2^(1/2)*s+2^(1/2)*s^3)
char=
160+s^4+(16+2^(1/2))*s^3+(8*2^(1/2)+68)*s^2+(16*2^(1/2)+144)*s
>>eval(char)
ans=
160+s^4+4901665356903357/281474976710656*s^3+2790603031573437/35184372088832*s^2+293
1340519928765/17592186044416*s
>>sym2poly(ans)
ans=
1.000017.414279.3137166.6274160.0000
HencetheCharacteristicequationis:
= s 4 + 17.4142 s 3 + 79.3137 s 2 + 166.6274 s + 160
USEACSYSRouthHurwitztoolasdescribedinpreviousproblemsandthisChapterssection214.
RH=
[1,5581205465083989*2^(46),160]
[87071/5000,5862680441794645*2^(45),0]
[427334336632381556219/6127076924293382144,160,0]
299
[238083438912827127943602680401244833403/1879436288300987963959490983755776000,
0,0]
[160,0,0]
Thefirstcolumnisallpositive,andthesystemisSTABLE.
Fortheothersection
symss
G=(s+1)/(s*(s+2)*(s^2+2*s+2))
g=ilaplace(G)
G=
(s+1)/s/(s+2)/(s^2+2*s+2)
g=
1/41/2*exp(t)*cos(t)+1/4*exp(2*t)
cltG=G/(1+G)
cltG=
(s+1)/s/(s+2)/(s^2+2*s+2)/(1+(s+1)/s/(s+2)/(s^2+2*s+2))
cltGsimp=simplify(cltG)
cltGsimp=
(s+1)/(s^4+4*s^3+6*s^2+5*s+1)
NexttypethedenominatorintoACSYSRouthHurwitzprogram.
2100
RH=
[1,6,1]
[4,5,0]
[19/4,1,0]
[79/19,0,0]
[1,0,0]
STABLE
2101
Chapter 3__________________________________________________________________________
3-1) a)
b)
c)
G H
d) Feedback ratio =
e)
3-2)
Characteristic equation: 1 2 1 0
3 2 1 0
31
3-3)
G 1
1 G1 H 1
H2
G2
G1
1 G1 H 1
G1G2 H2
G3 +
1 G1 H 1 G2
1
1 1
1
32
3-4)
G2
1 + G 2 G3 H 3
+ G2
X G1 G3 Y
1 + G2G3 H 3 + G2 H 2
-
H1
33
A Chapter3Solution
ns Golnarraghi,Kuo
+ G1G2 G3
X Y
1 + G 2 G3 H 3 + G 2 H 2
-
H1
G3
3
3-5)
34
3-6) MATLAB
symss
G=[2/(s*(s+2)),10;5/s,1/(s+1)]
H=[1,0;0,1]
A=eye(2)+G*H
B=inv(A)
Clp=simplify(B*G)
G=
[ 2/s/(s+2), 10]
[ 5/s, 1/(s+1)]
H=
1 0
0 1
A=
[ 1+2/s/(s+2), 10]
[ 5/s, 1+1/(s+1)]
B=
[ s*(s+2)/(s^2-48*s-48), -10/(s^2-48*s-48)*(s+1)*s]
[ -5/(s^2-48*s-48)*(s+1), (s^2+2*s+2)*(s+1)/(s+2)/(s^2-48*s-48)]
Clp =
[ -2*(24+25*s)/(s^2-48*s-48), 10/(s^2-48*s-48)*(s+1)*s]
[ 5/(s^2-48*s-48)*(s+1), -(49*s^2+148*s+98)/(s+2)/(s^2-48*s-48)]
35
A Chapter3Solution
ns Golnarraghi,Kuo
3-7)
33-8)
36
A Chapter3Solution
ns Golnarraghi,Kuo
3-9)
3
3-10)
3
3-11)
37
A Chapter3Solution
ns Golnarraghi,Kuo
3-12)
3-13)
symst
f=100*(10.33*exp(6*t)0
0.7*exp(10*t))
F=laplace(f)
symss
F=eval(F)
Gc=F*s
M=30000
symsK
Olp=simplify(KK*Gc/M/s)
Kt=0.15
Clp=simplify(Olp/(1+Olp*KKt))
s=0
Ess=eval(Clp)
f=
10030*exp(6 6*t)70*exp(10*t)
F=
80*(11*s+75)/s/(s+6)/(s+110)
ans=
38
A Chapter3Solution
ns Golnarraghi,Kuo
(880*s+6000)/s/(s+6)/(s+110)
Gc=
(880*s+6000)/(s+6)/(s+100)
M=
30000
Olp=
1/375*K*(11**s+75)/s/(s+66)/(s+10)
Kt=
0.1500
Clp=
2
20/3*K*(11*s s+75)/(2500*s^3+40000*ss^2+150000*ss+11*K*s+75*K)
s
s=
0
E
Ess=
2
20/3
3
3-14)
39
3-15)
1 1
Note: If G(s)=g(t),then {easG(s)}=u(ta)g(ta)
symsts
f=100*(10.3*exp(6*(t0.5)))
F=laplace(f)*exp(0.5*s)
F=eval(F)
Gc=F*s
M=30000
symsK
Olp=simplify(K*Gc/M/s)
Kt=0.15
Clp=simplify(Olp/(1+Olp*Kt))
s=0
Ess=eval(Clp)
digits(2)
Fsimp=simplify(expand(vpa(F)))
Gcsimp=simplify(expand(vpa(Gc)))
Olpsimp=simplify(expand(vpa(Olp)))
Clpsimp=simplify(expand(vpa(Clp)))
f=
10030*exp(6*t+3)
F=
(100/s30*exp(3)/(s+6))*exp(1/2*s)
F=
(100/s2650113767660283/4398046511104/(s+6))*exp(1/2*s)
Gc=
(100/s2650113767660283/4398046511104/(s+6))*exp(1/2*s)*s
M=
30000
Olp=
1/131941395333120000*K*(2210309116549883*s2638827906662400)/s/(s+6)*exp(1/2*s)
Kt=
0.1500
Clp=
310
A Chapter3Solution
ns Golnarraghi,Kuo
2
20/3*K*(2210 03091165498 883*s263882 279066624000)*exp(1/2*s)/(87960930 022208000000*s^2
5
52776558133 324800000*s+ +2210309116 6549883*K*eexp(1/2*s)*s2638827906 6662400*K*exxp(1/2*s))
s=
0
Ess=
2
20/3
Fsimp=
.10e3*exp(.550*s)*(5.*s6
6.)/s/(s+6.)
Gcsimp=
.10e3*exp(.550*s)*(5.*s6
6.)/(s+6.)
Olpsimp=
.10e2*K*exp p(.50*s)*(17
7.*s20.)/s/(s++6.)
Clpsimp=
5
5.*K*exp(.50
0*s)*(15.*s117.)/(.44e4*ss^2.26e5*s+1
11.*K*exp(.550*s)*s13.*K
K*exp(.50*s)))
3-16)
311
3-17)
3-18)
1
5 6 3 0.5
1
0.5
1
0.5 1.5 0.5 0.5 0.5
0.5 0.5
0.5 0.5
0.5
u1 0.5 z1
1
3
0.5
1/s -1 1/s 1/s x1 0.5
-6
0.5 0.5 x3 x2 0.5
1 -5
1.5 1
u2 0.5 0.5
-0.5 z2
312
3-19)
B1
B0 1/s 1 1/s
u y
x
-A1
-A0
3-20)
313
A Chapter3Solution
ns Golnarraghi,Kuo
3
3-21)
3
3-22)
314
A Chapter3Solution
ns Golnarraghi,Kuo
33-23)
315
A Chapter3Solution
ns Golnarraghi,Kuo
3-24)
3
3-25)
3
3-26)
3
3-27)
316
A Chapter3Solution
ns Golnarraghi,Kuo
33-28)
317
A Chapter3Solution
ns Golnarraghi,Kuo
33-29)
318
A Chapter3Solution
ns Golnarraghi,Kuo
3-30) Use Masonss formula:
3-31) MATLAB
symssK
G=100//(s+1)/(s+5)
g=ilaplace(G/s)
H=K/s
YN=simmplify(G/(1+G*H))
Yn=ilapplace(YN/s)

G=
100/(ss+1)/(s+5)

g=
25*exxp(t)+5*expp(5*t)+20

H=
K/s
319
YN=

100*s/(s^3+6*s^2+5*s+100*K)

ApplyRouthHurwitzwithinSymbolictoolofACSYS(seechapter3)

RH=

[1,5]
[6,100*k]
[50/3*k+5,0]
[100*k,0]
Stabilityrequires:0<k<3/10.
320
A Chapter3Solution
ns Golnarraghi,Kuo
332)
321
A Chapter3Solution
ns Golnarraghi,Kuo
333)M
MATLABsolu
utionsarein
n334.

334)M
MATLAB
clear all
p = [1
1 5 6 10] % Define polynomial
p s^3+5*s^2
2+6*s+10=0
roots(
(p)
G=tf(1
1,p)
step(G
G)
322
p=
15610

ans=
4.1337
0.4331+1.4938i
0.43311.4938i
Transferfunction:
1

s^3+5s^2+6s+10

323
Alternatively:
clear all
syms s
G=1/( s^3 + 5*s^2 + 6*s + 10)
y=ilaplace(G/s)
s=0
yfv=eval(G)
G =
1/(s^3+5*s^2+6*s+10)
y =
1/10+1/5660*sum((39*_alpha^2-91+160*_alpha)*exp(_alpha*t),_alpha =
RootOf(_Z^3+5*_Z^2+6*_Z+10))
s =
0
yfv =
0.1000

ProblemfindingtheinverseLaplace.
UseToolbox251tofindthepartialfractionstobetterfindinverseLaplace
clear all
B=[1]
A = [1 5 6 10 0] % Define polynomial s*(s^3+5*s^2+6*s+10)=0
[r,p,k]=residue(B,A)
B =
1
A =
1 5 6 10 0
r =
-0.0152
-0.0424 + 0.0333i
-0.0424 - 0.0333i
0.1000
p =
-4.1337
-0.4331 + 1.4938i
-0.4331 - 1.4938i
0
k =
[]
1 0.0152 0.0424 + 0.0333i 0.0424 - 0.0333i
SopartialfractionofYis: + + +
s s 4.1337 s 0.4331 + 1.4938i s 0.4331 - 1.4938i
324
A Chapter3Solution
ns Golnarraghi,Kuo
335)M
MATLABsolu
utionsarein
n336.
325
336)
clear all
p = [1 4 3 5 1] % Define polynomial s^4+4*s^3+3*s^2+5*s+1=0
roots(p)
G=tf(1,p)
step(G)
p =
1 4 3 5 1
ans =
-3.5286
-0.1251 + 1.1250i
-0.1251 - 1.1250i
-0.2212
Transfer function:
1
-----------------------------
s^4 + 4 s^3 + 3 s^2 + 5 s + 1
326
Alternatively:
clear all
syms s t
G=1/(s^4+4*s^3+3*s^2+5*s+1)
y=ilaplace(G/s)
s=0
yfv=eval(G)
G =
1/(s^4+4*s^3+3*s^2+5*s+1)
y =
1-
1/14863*sum((3955*_alpha^3+16873+14656*_alpha^2+7281*_alpha)*exp(_alpha*t),_alp
ha = RootOf(_Z^4+4*_Z^3+3*_Z^2+5*_Z+1))
s =
0
yfv =
1
ProblemfindingtheinverseLaplace.
UseToolbox251tofindthepartialfractionstobetterfindinverseLaplace
clear all
B=[1]
A = [1 4 3 5 1] % Define polynomial s^4+4*s^3+3*s^2+5*s+1=0
B =
1
A =
1 4 3 5 1
r =
-0.0235
-0.1068 + 0.0255i
-0.1068 - 0.0255i
0.2372
p =
-3.5286
-0.1251 + 1.1250i
-0.1251 - 1.1250i
-0.2212
k =
[]
327
A Chapter3Solution
ns Golnarraghi,Kuo
3-37)
3-38) MATLAB
M
Use TFFcal in ACSY
YS (go to ACSYS foldeer and type in
n TFcal in th
he MATLAB
B Command
d
Windoow).
TFcal
328
Alternatively use toolboxes 3-3-1 and 3-3-2

clear all
syms s
G1=100
G2=(s+1)/(s+2)
G3=10/s/(s+20)
G4=(101*s^3+2122*s^2+2040*s)/10/(s+1)
H1=1
H2=1
simplify(G1*G2*G3/(1+G1*G2*H1+G1*G2*H2+G1*G2*G3))
G1 =
100
G2 =
(s+1)/(s+2)
G3 =
10/s/(s+20)
G4 =
(101/10*s^3+1061/5*s^2+204*s)/(s+1)
H1 =
1
H2 =
1
ans =
1000*(s+1)/(201*s^3+4222*s^2+5040*s+1000)
clear all
TF=tf([1000 1000],[201 4222 5040 1000])
step(TF)
329
3-39)
clear all
syms s
P1 = 2*s^6+9*s^5+15*s^4+25*s^3+25*s^2+14*s+6 % Define polynomial
P2 = s^6+8*s^5+23*s^4+36*s^3+38*s^2+28*s+16 % Define polynomial
solve(P1, s)
solve(P2, s)
collect(P2-P1)
collect(P2+P1)
collect((P1-P2)*P1)
P1 =
2*s^6+9*s^5+15*s^4+25*s^3+25*s^2+14*s+6
P2 =
s^6+8*s^5+23*s^4+36*s^3+38*s^2+28*s+16
ans =
330
-1
-3
i*2^(1/2)
-i*2^(1/2)
-1/4+1/4*i*7^(1/2)
-1/4-1/4*i*7^(1/2)
ans =
-2
-4
i
-i
-1+i
-1-i
ans =
-s^6-s^5+8*s^4+11*s^3+13*s^2+14*s+10
ans =
3*s^6+17*s^5+38*s^4+61*s^3+63*s^2+42*s+22
ans =
-60+2*s^12+11*s^11+8*s^10-54*s^9-195*s^8-471*s^7-796*s^6-1006*s^5-1027*s^4-
848*s^3-524*s^2-224*s
Alternative:
clear all
P1 = [2 9 15 25 25 14 6] % Define polynomial
roots(P1)
P2 = [1 8 23 36 38 28 16] % Define polynomial
roots(P2)
P1 =
2 9 15 25 25 14 6
ans =
-3.0000
-0.0000 + 1.4142i
-0.0000 - 1.4142i
-1.0000
-0.2500 + 0.6614i
-0.2500 - 0.6614i
331
P2 =
1 8 23 36 38 28 16
ans =
-4.0000
-2.0000
-1.0000 + 1.0000i
-1.0000 - 1.0000i
0.0000 + 1.0000i
0.0000 - 1.0000i
3-40)
clear all
syms s
P6 = (s+1)*(s^2+2)*(s+3)*(2*s^2+s+1) % Define polynomial
P7 = (s^2+1)*(s+2)*(s+4)*(s^2+s+1) % Define polynomial
digits(2)
vpa(solve(P6, s))
vpa(solve(P7, s))
collect(P6)
collect(P7)
P6 =
(s+3)*(s+1)*(2*s^2+s+1)*(s^2+2)
P7 =
(s^2+1)*(s+2)*(s+4)*(s^2+s+1)
ans = -1.
-3.
1.4*i
-1.4*i
-.25+.65*i
-.25-.65*i
ans = -2.
-4.
i
-1.*i
-.50+.85*i
-.50-.85*i
ans =
2*s^6+9*s^5+15*s^4+25*s^3+25*s^2+14*s+6
ans =
8+s^6+7*s^5+16*s^4+21*s^3+23*s^2+14*s
332
341)
UseToolbox251tofindthepartialfractions
clear all
B= conv(conv(conv([1 1],[1 0 2]),[1 4]),[1 10])
A= conv(conv(conv([1 0],[1 2]),[1 2 5]),[2 1 4])
B =
1 15 56 70 108 80
A =
2 9 26 45 46 40 0
r =
-1.0600 - 1.7467i
-1.0600 + 1.7467i
0.9600
-0.1700 + 0.7262i
-0.1700 - 0.7262i
2.0000
p =
-1.0000 + 2.0000i
-1.0000 - 2.0000i
-2.0000
-0.2500 + 1.3919i
-0.2500 - 1.3919i
0
k =
[]
333
342)Usetoolbox332

clear all
B= conv(conv(conv([1 1],[1 0 2]),[1 4]),[1 10])
A= conv(conv(conv([1 0],[1 2]),[1 2 5]),[2 1 4])
G1=tf(B,A)
YR1=G1/(1+G1)
pole(YR1)
B=
115567010880
A=
29264546400
Transferfunction:
s^5+15s^4+56s^3+70s^2+108s+80
2s^6+9s^5+26s^4+45s^3+46s^2+40s
Transferfunction:
2s^11+39s^10+273s^9+1079s^8+3023s^7+6202s^6+9854s^5+12400s^4
+11368s^3+8000s^2+3200s
4s^12+38s^11+224s^10+921s^9+2749s^8+6351s^7+11339s^6+16074s^5
+18116s^4+15048s^3+9600s^2+3200s
ans=
0
0.7852+3.2346i
0.78523.2346i
2.5822
1.0000+2.0000i
1.00002.0000i
2.0000
0.0340+1.3390i
0.03401.3390i
0.2500+1.3919i
0.25001.3919i
0.7794
C= [1 12 47 60]
D= [4 28 83 135 126 62 12]
G2=tf(D,C)
YR2=G2/(1+G2)
pole(YR2)
C =
1 12 47 60
334
D =
4 28 83 135 126 62 12
Transfer function:
4 s^6 + 28 s^5 + 83 s^4 + 135 s^3 + 126 s^2 + 62 s + 12
-------------------------------------------------------
s^3 + 12 s^2 + 47 s + 60
Transfer function:
4 s^9+76 s^8+607 s^7+2687 s^6+7327 s^5+12899 s^4+14778 s^3+10618 s^2

+ 4284 s + 720
-------------------------------------------------------------------------------
4 s^9+76 s^8+607 s^7+2688 s^6+7351 s^5+13137 s^4+16026 s^3+14267 s^2+9924
s+4320
ans =
-5.0000
-4.0000
0.0716 + 0.9974i
0.0716 - 0.9974i
-1.4265 + 1.3355i
-1.4265 - 1.3355i
-3.0000
-2.1451 + 0.3366i
-2.1451 - 0.3366i
3-43) Use Toolbox 3-3-1

G3=G1+G2
G4=G1-G2
G5=G4/G3
G6=G4/(G1*G2)
G3=G1+G2
G4=G1-G2
G5=G4/G3
G6=G4/(G1*G2)
Transfer function:
8 s^12 + 92 s^11 + 522 s^10 + 1925 s^9 + 5070 s^8 + 9978 s^7 + 15154 s^6 +
18427 s^5
+ 18778 s^4 + 16458 s^3 + 13268 s^2 + 10720 s +

4800
335
-------------------------------------------------------------------------------
2 s^9 + 33 s^8 + 228 s^7 + 900 s^6 + 2348 s^5 + 4267 s^4 + 5342 s^3 + 4640 s^2
+
2400 s
Transfer function:
-8 s^12 - 92 s^11 - 522 s^10 - 1925 s^9 - 5068 s^8 - 9924 s^7 - 14588 s^6
- 15413 s^5 - 9818 s^4 - 406 s^3 + 7204 s^2 + 9760 s +

4800
-------------------------------------------------------------------------------
2 s^9 + 33 s^8 + 228 s^7 + 900 s^6 + 2348 s^5 + 4267 s^4 + 5342 s^3 + 4640 s^2
+
2400 s
Transfer function:
-16 s^21 - 448 s^20 - 5904 s^19 - 49252 s^18 - 294261 s^17 - 1.346e006 s^16
- 4.906e006 s^15 - 1.461e007 s^14 - 3.613e007 s^13 - 7.482e007 s^12
- 1.3e008 s^11 - 1.883e008 s^10 - 2.234e008 s^9 - 2.078e008 s^8 -

1.339e008 s^7
- 2.674e007 s^6 + 6.595e007 s^5 + 1.051e008 s^4 + 8.822e007 s^3 +

4.57e007 s^2
+
1.152e007 s
-------------------------------------------------------------------------------
16 s^21 + 448 s^20 + 5904 s^19 + 49252 s^18 + 294265 s^17 + 1.346e006 s^16
+ 4.909e006 s^15 + 1.465e007 s^14 + 3.643e007 s^13 + 7.648e007 s^12
+ 1.369e008 s^11 + 2.105e008 s^10 + 2.803e008 s^9 + 3.26e008 s^8 +

3.343e008 s^7
+ 3.054e008 s^6 + 2.493e008 s^5 + 1.788e008 s^4 + 1.072e008 s^3 +

4.8e007 s^2
336
+
1.152e007 s
Transfer function:
-16 s^21 - 448 s^20 - 5904 s^19 - 49252 s^18 - 294261 s^17 - 1.346e006 s^16
- 4.906e006 s^15 - 1.461e007 s^14 - 3.613e007 s^13 - 7.482e007 s^12
- 1.3e008 s^11 - 1.883e008 s^10 - 2.234e008 s^9 - 2.078e008 s^8 -

1.339e008 s^7
- 2.674e007 s^6 + 6.595e007 s^5 + 1.051e008 s^4 + 8.822e007 s^3 +

4.57e007 s^2
+
1.152e007 s
-------------------------------------------------------------------------------
8 s^20 + 308 s^19 + 5270 s^18 + 54111 s^17 + 379254 s^16 + 1.955e006 s^15
+ 7.778e006 s^14 + 2.471e007 s^13 + 6.416e007 s^12 + 1.383e008 s^11
+ 2.504e008 s^10 + 3.822e008 s^9 + 4.919e008 s^8 + 5.305e008 s^7 +

4.73e008 s^6
+ 3.404e008 s^5 + 1.899e008 s^4 + 7.643e007 s^3 + 1.947e007 s^2 +

2.304e006 s
337
Chapter 4__________________________________________________________________________
4-1) When the mass is added to spring, then the spring will stretch from position O to position L.
The total potential energy is:
1
2
where y is a displacement from equilibrium position L.
The gravitational energy is:
The kinetic energy of the mass-spring system is calculated by:
1
2
As we know that , then
By differentiating from above equation, we have:
0
since KL = mg, therefore:
0
As y cannot be zero because of vibration, then 0
41
4-2)
KU
KL
4-3) a) Rotational kinetic energy:
Translational kinetic energy:
Relation between translational displacement and rotational displacement:
1
2
Potential energy:
As we know , then:
42
1 1 1
2 2 2
By differentiating, we have:
0
Since cannot be zero, then 0
b)
c)
1 1 1
2 2 2
where at the maximum energy.
1 1
2 2
Then:
1 1
2 2
Or:
43
J
d) G (s) =
(ms + K )
2
% select values of m, J and K

%Step input
K=10;
J=10;
M=1;
G=tf([10],[1 0 10])
step(G,10)
xlabel( 'Time(sec)');
ylabel('Amplitude');
Transferfunction:
10
s^2+10
44
AutomaticConttrolSystems,9thEdition Chapteer4Solutionss Golnarraghi,Kuo
4-44)
(a)Forceequaations:
dy1 dy2 + K y y
2
d y1 dy1
f (t ) = M 1 + B1 + B3 ( 1 2)
dt
2
dt dt dt

d 2
2
dy1 dy d y2 dy2
B3 + K ( y1 y2 ) + M 2 2 + B2
dt ddt dt dt
Rearraangetheequattionsasfollows:
2
d y1 (B + B3 ) dy1 B3 dyy 2 K f
2
= 1
+ (y 1
y2 ) +
dt M1 dt M 1 dt
d M1 M1

2
d y2 B3 dy1 (B + B3 ) dy2 K
2
= 2
+ (y1
y2 )
dt M 2 dt M2 dtt M2
(i)Staatediagram:SSince y1 y 2 appearsasone
a unit,theminimumnumberofintegratorsisthree.
45
dy 2 dy1
Stateequations:Definethestatevariablesas x1 = y1 y 2 , x 2 = , x3 = .
dt dt
dx1 dx2 K (B + B3 ) B3 dx3 K B3 (B + B3 ) 1

= x2 + x3 = x1 2
x2 + x3 = x1 + x2 1
x3 + f
dt dt M2 M2 M2 dt M1 M1 M1 M
dy 2 dy1
(ii)Statevariables: x1 = y 2 , x 2 = , x3 = y1 , x 4 = .
dt dt
Stateequations:
dx1 dx 2 K B2 + B3 K B3
= x2 = x1 x2 + x3 + x4
dt dt M2 M2 M2 M2
dx3 dx 4 K B3 K B1 + B3 1
= x4 = x1 + x2 x3 x4 + f
dt dt M1 M1 M1 M1 M1
Statediagram:
Transferfunctions:
M 2 s + ( B2 + B3 ) s + K
2
Y1 ( s )
=
F (s) {
s M 1 M 2 s + [( B1 + B3 ) M 2 + ( B2 + B3 ) M 1 ] s + [ K ( M 1 + M 2 ) + B1 B2 + B2 B3 + B1 B3 ] s + ( B1 + B2 ) K
3 2
}

Y2 ( s ) B3 s + K
=
F (s) {
s M 1 M 2 s + [ ( B1 + B3 ) M 2 + ( B2 + B3 ) M 1 ] s + [ K ( M 1 + M 2 ) + B1 B2 + B2 B3 + B1 B3 ] s + ( B1 + B2 ) K
3 2
}
(b)Forceequations:
46
2
d y1 (B + B2 ) dy1 B2 dy 2 1 dy2 dy1 K
2
= 1
+ + f = y2
dt M dt M dt M dt dt B2
(i)Statediagram:

Definetheoutputs
oftheintegrators
asstatevariables,
dy
x1 = y 2 , x 2 = 1
dt
.

Stateequations:
dx1 K dx 2 K B1 1
= x1 + x 2 = x1 x2 + f
dt B2 dt M M M
dy1
(ii)Stateequations:Statevariables: x1 = y 2 , x 2 = y1 , x3 = .
dt
dx1 K dx 2 dx3 K B1 1
= x1 + x3 = x3 = x1 x3 + f
dt B2 dt dt M M M
Transferfunctions:
Y1 ( s) B2 s + K Y2 ( s ) B2
= =
F ( s ) s MB2 s + ( B1 B2 + KM ) s + ( B1 + B2 ) K
2
F (s) MB2 s + ( B1 B2 + KM ) s + ( B1 + B2 ) K
2
(c)Forceequations:
dy1 dy2 1
2
d y2 (B + B2 ) dy2 B1 dy2 B1 dy1 K
= + f 2
= 1
+ + y2
dt dt B1 dt M dt M dt M dt M
(i)Statediagram:
47
Stateequatio
ons:Definetheeoutputsofintegratorsassttatevariables.
dx1 dxx 2 K B2 1
= x2 = x1 x2 + f
dt d
dt M M M
dy 2
(ii)Stateequations:statevariaables: x1 = y1 , x 2 = y 2 , x3 = .
dt
dx1 1 dx 2 dxx3 K B2 1
= x3 + f = x3 = x2 x3 + f
dt B1 dt d
dt M M M
Statediagram
m:
Transferfuncctions:
Ms + ( B1 + B2 ) s + K
2
Y1 ( s ) Y2 ( s ) 1
= =
F (s) (
B1 s Ms + B2 s + K
2
) F (s) Ms + B2 s + K
2
4-55)
48
(a)Forceequaations:
2
1 d y2 B dy 2 K1 + K 2 K2
y1 = ( f + Mg ) + y 2 2
= y2 + y1
K2 dt M dt M M
Staatediagram:

Staateequations:
Deffinethestate
varriablesas:
dy
x1 = y 2 , x 2 = 2
d
dt
.
dx1 d 2
dx K1 B 1
= x2 = x1 x2 + ( f + Mg )
dtt dt M M M
Transferfunction
ns:
2
Y1 ( s ) s + Bs + K1 + K 2 Y2 ( s ) 1
= 2
= 2

F ( s) K 2 ( Ms + Bs + K1 ) F ( s) Mss + Bs + K1
(b)Forcceequations:
dy1 dy2 + K1 y y B2 y y B2 dy2

2
dy1 1 dy 2 K1 d y2 B1
= [ f (t ) + Mg ] + (y y2 ) = ( 1 2) ( 1 2)
M dt dt M
1 2
dt B1 dt B1 dt M M dt
Stattediagram:(W
Withminimumnumberofinttegrators)
49
ToobtainthetranssferfunctionsY1 ( s ) / F ( s ) andY2 ( s ) / F ( s ), weneedtoredefinethesttatevariablesaas:
x1 = y 2 , x 2 = dy 2 / dtt , and x3 = y1 .
Stattediagram:
Transferfuncctions:
Ms + ( B1 + B2 ) s + K1 Bs + K1
2
Y1 ( s ) Y2 ( s )
= =
F (s) s
2
[ MB
M s + (B B
1 1 2
+ MK1 )] F (s) s
2
[ MB
M s + (B B
1 1 2
+ MK1 )]
4-66) a)
y1
y2
K ( y1 y 2 )
m K ( y1 y 2 ) M F
mgy& 2 Mgy&1
b) From Newttons Law:
If y1 and y2 are considerred as a posiition and v1 and

a v2 as vellocity variabbles
Then:
The output equation cann be the veloocity of the engine,

e whicch means
410
c)
Obtaining requires solving above equation with respect to Y2(s)
From the first equation:
Substituting into the second equation:
By solving above equation:
1
2
d)
0 0 1 0
0 0 0 1 0
0
0 1
0 0
0001 0
4-7)(a)Forceequations:
dy1 dy2 dy1 dy2 = M d y2 + B dy2

2
f (t ) = K h ( y1 y2 ) + Bh K h ( y1 y 2 ) + Bh
dt dt dt dt 2 t
dt dt
dy 2
(b)Statevariables: x1 = y1 y 2 , x 2 =
dt
411
Stateequations:
dx1 Kh 1 dx 2 Bt 1
= x1 + f (t ) = x2 + f (t )
dt Bh Bh dt M M
4-8)
2
1
For the left pendulum:
sin
sin sin cos sin sin cos
0
sin sin sin cos
For the right pendulum, we can write the same equation:
sin sin sin cos
since the angles are small:
sin
cos 1 0

sin 0
cos 1
4-9) a)
412
Fy
m
x(t) Fx
mg
M
f

b) If we consider the coordinate of centre of gravity of mass m as (xg, yg),
Then sin and cos
From force balance, we have:
cos sin
From the torque balance, we have:
cos sin sin
Where:
cos
cos sin
Substituting these equation:
cos sin
4-10) a)
413
Fy1 Fy 2
Fx1 Fx 2
m1 g
m2 g
1 2
b) Kinetic energy
(i) For lower pendulum:
1 1
sin cos
2 2
For upper pendulum:
1 1
sin cos
2 2
For the cart:
(ii) Potential energy:
For lower pendulum: cos
For upper pendulum: cos
For the cart: 0
(iii) Total kinetic energy:
Total potential energy:
414
The Lagrangian equation of motion is:
Substituting T and U into the Lagrangian equation of motion gives:
cos cos sin sin

cos sin
cos sin
4-11) a) From the Lagrangian equation of motion:
sin 0
b) As:
Then
sin 0
If we linearize the equation about beam angle = 0, then sin and sin
Then:
415
c) Considering
Then the state-space equation is described as:
0
0 1
0 0
mgd
d) G ( s) =
( s L( J / r 2 + m))
2
clear all
% select values of m, d, r, and J
%Step input
g=10;
J=10;
M=1;
D=0.5;
R=1;
L=5;
G=tf([M*g*D],[L*(J/R^2+M) 0 0])
step(G,10)
Transfer function:
5
------
55 s^2
416
4-12) If the aircraft is at a constant altitude and velocity, and also the change in pitch angle does not
change the speed, then from longitudinal equation, the motion in vertical plane can be written as:
sin
cos
Where u is axial velocity, is vertical velocity, q is pitch rate, and is pitch angle.
Converting the Cartesian components with polar inertial components and replace x, y, z by T, D,
and L. Then we have:
1
cos sin
1
sin cos
Where = is an attack angle, V is velocity, and is flight path angle.
417
It should be mentioned that T, D, L and M are function of variables and V.
Refer to the aircraft dynamics textbooks, the state equations can be written as:
b) The Laplace transform of the system is:
By using Laplace transform, we have:
1
2
3
From equation (1):
Substituting in equation (2) and solving for q(s):
Substituting above expression in equation (3) gives:
If we consider sin , then
By using Laplace transform:
Which gives:
418
For plotting
g state flow diagram, equation
u (4) muust be rewritten as:
or
S the state flow diagram

So, m will plotteed as:
ml 1/Ms 1 1/s
u y
x
-B/M
-K
Also look at seection 4-11
4-113)(a)Torqueequation: Stated
diagram:
2
d B d 1
2
= + T (t )
dt J dt J
Stateequations:
dx1 dx 2 B 1
= x2 = x2 + T
dt dt J J
Transsferfunction:
( s ) 1
=
T ( s) s( Js + B )
(b)Torqueequations:
d 1 d 2
2
K 1
2
= ( 1
2 ) + T K ( 1 2 ) = B
dt J J dt
Statediagram:(minimumnumbeerofintegratorrs)
419
Stateequations:
dx1 K dx 2 K 1
= x1 + x 2 = x1 + T
dt B dt J J
d 1
Stateequations:Let x1 = 2 , x 2 = 1 , and x3 = .
dt
dx1 K K dx 2 dx3 K K 1
= x1 + x2 = x3 = x1 x2 + T
dt B B dt dt J J J
Statediagram:
Tran
nsferfunctionss:
1 ( s ) B +K
Bs 2 ( s) K
= =
T (s) (
s BJs + JKs + BK
2
) T (s) (
s BJs + JK
2
Ks + BK )
(c)Torqqueequationss:
d 1 d 2
2 2
T (t ) = J 1 2
+ K ( 1 2 ) K ( 1 2 ) = J 2 2
dt dt
Stattediagram:
420
d 2 d 1
eequations:sttatevariables:: x1 = 2 , x 2 =
State , x3 = 1 , x 4 = .
dt d
dt
dx1 dx 2 K K d 3
dx d 4
dx K K 1
= x2 = x1 + x3 = x4 = x1 x3 + T
dt dt J2 J2 dt dt J1 J1 J1
nsferfunctionss:
Tran
1 ( s ) J2s + K 2 ( s)
2
K
= =
s J 1 J 2 s + K ( J 1 + J 2 ) s J 1 J 2 s + K ( J 1 + J 2 )
2 2 2 2
T (s) T (s)
(d)Torqueequationss:
d m d 1 d 2
2 2 2
T (t ) = J m 2
+ K1 ( m 1 ) + K 2 ( m 2 ) K1 ( m 1 ) = J 1 2
K 2 ( m 2 ) = J 2 2

dt dt dt
421
Statediiagram:

d 1 d m d 2
ons: x1 = m 1 , x 2 =
Stateequatio , x3 = , x 4 = m 2 , x5 = .
dtt dt dt
dx1 dx 2 K1 dx3 K1 K2 1 dx 4 dx5 K2

= x 2 + x3 = x1 = x1 x4 + T = x3 x5 = x4
dt dt J1 dt Jm Jm Jm dt dtt J2
Transferfuncctions:
1 ( s ) K1 ( J 2 s + K 2 )
2
=
s s + ( K1 J 2 J m + K 2 J 1 J m + K1 J 1 J 2 + K 2 J 1 J 2 ) s + K1 K 2 ( J m + J 1 + J 2 )
2 4 2
T (s)

2 (s) K 2 ( J 1 s + K1 )
2
=
s s + ( K1 J 2 J m + K 2 J 1 J m + K1 J 1 J 2 + K 2 J 1 J 2 ) s + K1 K 2 ( J m + J 1 + J 2 )
2 4 2
T (s)
(e)Torqqueequationss:
d 2 m K1 K2 1 d 1
2
K1 B1 d 1 d 2
2
K2 B2 d 2
2
= ( m 1 ) ( m 2 ) + T = ( m
1 ) 2
= ( m
1 )
dt Jm Jm Jm d
dt
2
J1 J 1 dt dt J2 J 2 dt
Statediagram
m:
422
d 1 d m d 2
SStatevariabless: x1 = m 1 , x 2 = , x3 = , x 4 = m 2 , x5 = .
dt dt dt
Stateequatio
ons:
dx1 dx 2 K1 B dx3 K1 K2 1 dx4 dx5 K2 B2
= x 2 + x3 = x1 1 x 2 = x1 x4 + T = x3 x 5 = x4 x5
dt dt J1 J1 dt Jm Jm Jm dt dt J2 J2
Transferfuncctions:

1 ( s )
=
(
K1 J 2 s + B2 s + K 2
2
) 2 (s)
=
(
K 2 J 1 s + B1 s + K1
2
)
T (s) (s) T (s) (s)
( s ) = s { J 1 J 2 J m s + J m ( B1 + B2 ) s + [( K1 J 2 + K 2 J 1 ) J m + ( K1 + K 2 ) J 1 J 2 + B1 B2 J m ] s
2 4 3 2

+ [( B1 K 2 + B2 K1 ) J m + B1 K 2 J 2 + B2 K1 J 1 ] s + K1 K 2 ( J m + J 1 + J 2 )}
4-114)
d 1 d 3
2 2
N1 N3 N1
Tm (t ) = J m 2
+ T1 T1 = T2 T3 = T4 T4 = J L 2
T2 = T3 2 = 1
dt N2 N4 dt N2
(a))
N1 N 3 N3 N3 d 3
2
d 1
2

N1 N 3 N1 N 3 d 2 1
2
3 = 1 T2 = T4 = JL Tm = J m + T4
= m
J + JL 2
N 2 N 4 dt
2 2
N2 N4 N4 N4 dt dt N2 N4
423
(b)
d 1 d 2 d 3
2 2 2
N1 N3
Tm = J m 2
+ T1 T2 = J 2 2
+ T3 T4 = ( J 3 + J L ) 2
T1 = T2 T3 = T4
dt dt dt N2 N4

d 2 d 2 d 3
2 2 2
N1 N1 N 3 N3 N3
2 = 1 3 = 1 T2 = J 2 2
+ T4 = J 2 2
+ (J 3
+ JL ) 2
N2 N2 N4 dt N4 dt N4 dt
d 1
2
N1 N1
d 2
2
2
N1 N 3
N3
2
d 21 d 3
2
Tm (t ) = J m 2 + J 2 2 + ( J 3 + J 4 ) 2 = J m + J 2 + ( J 3 + J L ) dt 2
dt N2 dt N4 dt N2 N2 N4
4-15)(a)
2 2
d m d L N1
Tm = J m 2
+ T1 T2 = J L 2
+ TL T1 = T2 = nT2 m N1 = L N 2
dt dt N2
d m d L J m + nJ + nT nTm n TL
2 2 2
Tm = J m + nJ L + nTL = L
Thus, L =
n
L L
Jm + n JL
2 2 2
dt dt
L J mTL Jm
Set
n
= 0. (Tm ( 2
)
2 nTL ) J m + n J L 2 nJ L nTm n J L = 0 Or, n + ( 2
) 2
n = 0
J LTm JL

2 2 2
J mTL J mTL + 4 J m J LTm
Optimalgearratio: n = + wherethe+signhasbeenchosen.
2 J LTm 2 J LTm
(b)When TL = 0 ,theoptimalgearratiois

n = Jm / J L
4-16)(a)Torqueequationaboutthemotorshaft: Relationbetweenlinearandrotationaldisplacements:
2 2
d m 2 d m d m
Tm = J m 2
+ Mr 2
+ Bm y = r m
dt dt dt
(b)TakingtheLaplacetransformoftheequationsinpart(a),withzeroinitialconditions,wehave
Tm ( s ) = J m + Mr ( 2
)s 2
m
( s ) + Bm s m ( s ) Y (s) = rm (s)
424
Transfferfunction:
Y ( s) r
=
Tm ( s ) (
s J m + Mr
r
)s + B m

4-117)(a)
d m
2
Tm = J m 2
+ r ( T1 T2 ) T1 = K 2 ( r m r p ) = K 2 ( r m y ) T2 = K1 ( y r m )
dt
d m
2 2 2
d y d y
T1 T2 = M 2
us, Tm = J m
Thu 2
+ r ( K1 + K 2 )( r m y ) M 2
= ( K1 + K 2 )( r m y )
dt dtt dt
(c)Stateequations:
dx1 dx2 K1 + K 2 dx3 r ( K1 + K 2 ) 1

= rx3 x2 = x1 = x1 + Tm
dt dt M dt Jm Jm
(d)Transferfunction:
Y ( s) r ( K1 + K 2 )
=
s J m Ms + ( K1 + K 2 )(( J m + rM )
2 2
Tm ( s )
(e)Charaacteristicequaation:
s
2
J m Ms 2 + ( K1 + K 2 )( J m + rM ) = 0
4-118) (a)Systeemequations:
425
d m dia
Tm = K i ia = ( J m + J L ) + Bm m ea = Ra ia + La + K b m y = n m y = y ( t TD )
dt dt
d
TD = (sec) e = r b b = Ks y E a ( s ) = KGc ( s ) E ( s )
V
Blockkdiagram:
(b)Forw
wardpathtransferfunction
n:
TD s
Y (s) KK i nGc ( s )e
=
E (s) s {( Ra + La s ) [( J m + J L ) s + Bm ] + K b K i }
Clossedlooptranssferfunction:
TD s
Y ( s) KK
K i nGc ( s )e
=
s ( Ra + La s ) [ ( J m + J L ) s + Bm ] + K b K i s + KGc ( s ) K i ne
TD s
R(s)
4-119) (a)Torquueequations:
d m d m d L d L
2 2
Tm (t ) = J m 2
+ Bm + K ( m L ) K ( m L ) = J L 2
+ BL
dt dt dt dt
Statediagram:
426
(b)Transsferfunctions:
L ( s) m (s) J L s + BL s + K
2
K
= = ( s ) = s J m J L s + ( Bm J L + BL J m ) s + ( KJ m + KJ L + Bm BL ) s + Bm K
3 2
Tm ( s ) (s) Tm ( s ) ( s)
(c)Charaacteristicequaation: ( s ) = 0
Tm
dystateperforrmance: Tm ( t ) = Tm = constant.
(d)Stead t Tm ( s ) = .
s
J L s + BL s + K
2
1
lim m (t ) = lim s m ( s ) = lim =
J m J L s + ( Bm J L + BL J m ) s + ( KJ
K m + KJ L + Bm BL ) s + Bm K
3 2
t s 0 s 0 Bm
Thuss,inthesteadyystate, m = L .
(e)Theesteadystatevvaluesof m and
a L donotdependon J m and J L .
4-220)(a)Torqqueequation:(AbouttheceenterofgravityyC)
2
d
J 2
= Ts d 2 sin + Fd d1 Fa d1 = J 1 = K F d1 sin
dt
2 2
d d
Thuss, J 2
= Ts d 2 + K F d1 J 2
K F d1 = Ts d 2
dt dt
2
(b) Js ( s ) K F d1 ( s ) = Ts d 2 ( s )
(c)With
hCandPintercchanged,thetorqueequationaboutCis:
d d
2 2
Ts ( d1 + d 2 ) + F d 2 = J 2
Ts ( d1 + d 2 ) + K F d 2 = J 2

dt dt
427
( s ) Ts ( d1 + d 2 )
Js ( s ) K F d 2 ( s ) = Ts ( d1 + d 2 ) ( s ) =
2

(s) Js K F d 2
2
4-21)(a)Nonlineardifferentialequations:
dx ( t ) dv ( t )
= v(t ) = k ( v ) g ( x ) + f ( t ) = Bv ( t ) + f ( t )
dt dt
e( t ) e( t 0
With Ra = 0 , ( t ) = = K f i f ( t ) = K f i f ( t ) = K f ia ( t ) Then, ia ( t ) =
Kb v ( t ) Kb K f v ( t )
2
Ki e ( t ) dv ( t ) Ki 2
f ( t ) = Ki ( t )ia ( t ) = 2 2
. Thus, = Bv ( t ) + 2 2
e (t )
Kb K f v (t ) dt Kb K f v (t )
(b)Stateequations: ia ( t ) asinput.
dx ( t ) dv ( t ) 2
= v (t ) = Bv ( t ) + Ki K f ia ( t )
dt dt
(c)Stateequations: ( t ) asinput.
2 (t )
f ( t ) = K i K f ia ( t ) ia ( t ) = i f ( t ) =
Kf

dx ( t ) dv ( t ) Ki 2
= v (t ) = Bv ( t ) + (t )
dt dt Kf
428
4-22) Define as the angle between mass m and the horizontal axis (positive in c..c.w. direction):
=t ym
Use Newtons second law:
y + &&
m( && ym ) = Fm
( M m) &&
y = Fm By& Ky
ym = e 2 sin t
&&
My&& + By& + Ky = me 2 sin t
Where M is the Mass of the overall block system. M-m is the mass of the block alone.
Y me 2
G ( s) = =
R Ms 2 + Bs + K
Zero i.c. and input r (t ) = sin t
Note = t . So in case of a step response as asked in the question, is a step input and angle increases
with time i.e. it is a ramp function. Hence, ym is a sinusoidal function, where the Laplace transform of a

sine function is sin(t ) = 2
s + 2
Pick values of the parameters and run MATLAB. See toolbox 5-8-2
clear all
m=20.5 %kg
M=60 %kg
K=100000 %N/m
Om=157 %rad/s
B=60 %N-m/s
e=0.15 %m
G=tf([m*e*Om^2],[M B K])
t=0:0.01:1;
u=1*sin(Om*t);
lsim(G,u,t)
429
m =
20.5000
M =
60
K =
100000
Om =
157
B =
60
e =
0.1500
Transfer function:
7.58e004
----------------------
60 s^2 + 60 s + 100000
430
4-23) a) summation of vertical forces gives:
2 1
0 2
If we consider and , then:
2
0
The state-space model is:
0 0 1 0
0 0 0 1 0
2 0
0 0 1
0 0 0
b)
By applying Laplace transform for equations (1) and (2), we obtain:
Which gives:
and
Therefore:
2 2
4-24) a) Summation of vertical forces gives:
Consider and , then
431
So, the state-space model of the system is:

0 0 1 0
0 0 0 1 0
0
1
b) The Laplace transform of the system is defined by:
where
as a result:
Substituting into above equation:
4-25) a) According to the circuit:
0
2 2
0
2 2
0
2 2
By using Laplace transform we have:
432
V s V s V s V s
CsV s 0
2R 2R
Cs V s Cs
V s V s V s V s 0
2 R 2
Cs V s V s
V s V s 0
2 2R
From above equations:
1
V s V s V s
2 RCS 1
RCS
V s V s V s
2 RCS 1
Substituting V1(s) and V2(s) into preceding equations, we obtain:
1
4 1
b) Measuring Vout requires a load resistor, which means:
i1 i2 RL
Vin VC1 VC 2
Then we have:
When
If RL >>R, then
433
By using Laplace transform we have:
Therefore:
can be obtained by substituting above expressions into the first equation of the state variables
of the system.
4-26) a) The charge q is related to the voltage across the plate:
The force fv produced by electric field is:
2
Since the electric force is opposes the motion of the plates, then the equation of the motion is written
as:
The equations for the electric circuit are:
As we know, and , then:
434
Since , then :
b) As then
If 1
Then the transfer function is:
435
If 1
Then the transfer function is:
4-27) a) The free body diagram is:
where F is required force for holding the core in the equilibrium point against magnetic
field
b) The current of inductor, i, and the force, F, are function of flux, , and displacement, x.
Also, we know that
436
The total magnetic field is:
,
2
where W is a function of electrical and mechanical power exerted to the inductor, so:
W
i
L x
W
F
x 2L x
As v = , then:
1
2
c) Changing the flux requires a sinusoidal movement, and then we can conclude that:
if the inductance is changing relatively, then L(x) = Lx, where L is constant.
Also, the current is changing with the rate of changes in displacement. It means:
i Bx
So:
Substituting these equations into the state-space equations gives:
1
2
Therefore:
V s 2
tan 1
F s B
437
4-28) a) The free body diagram is:
where F is the external force required for holding the plate in the equilibrium point against the
electrical field.
b) The voltage of capacitors, , and the force, , are function of charge, , and displacement, .
Also, we know
The total electrical force between plates is:
,
2
Where W is a function of electrical and mechanical power exerted to the capacitor, so:
As , then:
1
2
c) The same as Problem 4.28,
Consider:
Then solve the equations.
438
4-29) According to the circuit:

_ _
or
As an op-amp is modeled with the following equation:

_
1
Then:
1
1 1
1
4-30) a) Positive feedback ratio:
b) Negative feedback ratio:
c) According to the circuit:
Therefore:
439
As
10
1
then:
10
1
1
which gives:
1
10 1 10
It is stable when 1 10 0 which means: 10
4-31) a) If the drop voltage of Rin is called v1
Then:
Also:
0
Then:
Substituting this expression into the above equation gives:
1
1 0
As a result:
440
1 1
1
Or
1
b) If the dropped voltage across resistor Rf is called vf, then

0
0
As a result:
Substituting into the second equation gives:
or
As a result:
1
1
4-32) The heat flow-in changes with respect to the electric power as:
where R is the resistor of the heater.
The heat flow-out can be defined as:
where Kf is the heat flow coefficient between actuator and air, T1 and T2 are temperature of
actuator and ambient.
441
Since the temperature changes with the differences in heat flows:
1 1 1
where C is the thermal capacitor.
The displacement of actuator is changing proportionally with the temperature differences:
If we consider the T2 is a constant for using inside a room, then
Therefore:
By linearizing the right hand side of the equation around point

1
2 1
Or
2 1
If we consider the right hand side of the above equation as two inputs to the system as: 2
and 1 or , then:
2
1
4-33) Due to insulation, there is no heat flow through the walls. The heat flow through the sides is:
442
2
, 1
ln
2
, 2
ln
Where T1 and T2 are the temperature at the surface of each cylinder.
As , , , then from equation (1) and (2), we obtain:
ln
, 3
2
The conduction or convection at:
: 4
: 5
: 6
The thermal capacitance dynamics gives:
, 7
Where
According to the equation (7) and (8), T1 and Tf are state variables.
Substituting equation (3), (4), (5) and (6) into equation (7) and (8) gives the model of the system.
4-34) As heat transfer from power supply to enclosure by radiation and conduction, then:
2
1 1 1 1
443
Also the enclosure loses heat to the air through its top. So:
Where
And Ct is the convective heat transfer coefficient and At is the surface area of the enclosure.
The changes if the temperature of heat sink is supposed to be zero, then:
Therefore where , as a result:
According to the equations (1) and (4), Tp and Te are state variables. The state model of the system
is given by substituting equations (2), (3), and (6) into these equations give.
4-35) If the temperature of fluid B and A at the entrance and exit are supposed to be and , and
TAN and TAX, respectively. Then:
1
2
The thermal fluid capacitance gives:
From thermal conductivity:
5
ln
1 1
2
444
Where Ci and Co arre convectivve heat transffer coefficiennt of the inner and outerr tube; Ai andd Ao are
the surfface of innerr and outer tuube; Ri and Ro are the raadius of inneer and outer tube.
t
Substittuting equations (1), (2), and (5) intoo equations (3) and (4) giives the statee model of thhe
system
m.
4-336)(a)Blockkdiagram:
(b)Transferfunction:
D s
( s ) K1 K 4 e
=
Js + ( JK L + B ) s + K 2 B + K 3 K 4 e
D s
( s)
2
(c)Characteristicequattion:
Js + ( JK L + B ) s + K 2 B + K 3 K 4 e
D s
= 0
2

(d)Transferfunction:
( s ) K1 K 4 ( 2 D s )

(s) ( s)
Charaacteristicequaation:
( s ) J D s + ( 2 J + JK 2 D + B D ) s + ( 2 JK 2 + 2 B D K 2 B D K 3 K 4 ) s + 2 ( K 2 B + K 3 K 4 ) = 0
3 2

4-337) The tottal potential energy is:
1 1
2 2
The tottal kinetic en
nergy is:
445
Therefore:
2
As a result:
2
So, the natural frequency of the system is calculated by:
Also, by assuming sin and substituting into yields the same result
when calculated for maximum displacement.
4-38) If the height of the reservoir, the surge tank and the storage tank are assumed to be H, h1 and h2,
then potential energy of reservoir and storage tank are:
For the pipeline we have:
The surge tank dynamics can be written as:
At the turbine generator, we have:
where I is a known input and Q2-v is the fluid flow transfer between point 2 and valve. The
behaviour of the valve in this system can be written as:
446
| |
Regarding Newtons Law:
According to above equations, it is concluded that Q and h1 are state variables of the system.
The state equations can be rewritten by substituting P2, Pv, Ps and Q2-v from other equations.
4-39)
If the beam rotate around small angle of cos 1 , then
where A and E are cross sectional area and elasticity of the cable; H is the distance between point
O and the bottom of well, and y is the displacement.
On the other hand, Newtons Law gives:
447
Pa A
Bv 2
Pb A
where B is the viscous friction coefficient, Aw is the cross sectional area of the well; P1 and P2 are
pressures above and below the mass m.
The dynamic for the well can be written as two pipes separating by mass m:
Pb A
Ps A
Q
Q y
Ff
Pa A P1 A
a b)
0
8
0
Where D is the distance between point O and ground, Ps is the pressure at the surface and known.
If the diameter of the well is assumed to be r, the Ff for the laminar flow is
32
Therefore:
448
32
32
The state variables of the system are , v, y, Q, Q1.
4-40) For the hydraulic amplifier, we have:
As a result
where N is a constant and A is the cross sectional area.
For the walking beam:
For the spring:
The angular velocity of the lever is assumed as:
0 0
The moments of inertia of the lever are calculated as:
sin
cos sin
cos
0
cos sin
where L is the length of lever and r is the offset from the center of rotation.
According to the equation of angular motion:
449
Also:
sin cos
2
Due to force balance, we can write:
sin sin
2
Therefore can be calculated form above equations.
On the other hand, sin , and and , the dynamic of the system is:
where B is the viscous friction coefficient, and n1 and n2 are constant.
The state variables of the systems are ,yp, and 2.
450
4-41) If the capacitances of the tanks are assumed to be C1 and C2 respectively, then
Therefore:
1
Asa result:
1 1 1
0
1 1
0
4-42) The equation of motion is:
Considering gives:
or
4-43)(a)Blockdiagram:
451
(b)Transfferfunction:
TAO ( s ) KM KR 3.51
= =
Tr ( s ) (1 + s )((1 + s ) + K
c s m
KR 20 s + 122 s + 4.51
2
4-444)
Sysstemequations:
d m d m d L d L
2 2
Tm (t ) = J m 2
+ Bm + K ( m L ) K ( m L ) = J L 2
+ Bp
dt dt dt d
dt
E L
uation: eo =
Outputequ
20
Statediiagram:
Transferr
fun
nction:
L (s) K
=
s J m J L s + ( Bm J L + B p J m ) s + ( J m K + J L K + Bm B p ) s + Bm K
3 2
Tm ( s )
Eo ( s ) KE / 20
=
s J m J L s + ( Bm J L + B p J m ) s + ( J m K + J L K + Bm B p ) s + Bm K
3 2
Tm ( s )
4-445)(a)Statteequations:
d L d L K2 K2 d t d t K1 K1
=L = m L = t = m t
dt dtt JL JL dt dtt Jt Jt
452
d m d m Bm (K + K2 ) K1 K2 1
= m = m 1
m + t + L + Tm
dt dtt Jm Jm Jm Jm Jm
(b)Stateddiagram:
(c)Transfferfunctions:

L (s)
=
(
K 2 J t s + K1
2
) t ( s)
=
(
K1 J L s + K 2
2
) m (s)
=
J t J L s + ( K1 J L + K 2 J t ) s + K1 K 2
4 2

Tm ( s ) (s) Tm ( s ) (s) Tm ( s ) (s)
( s ) = s[ J m J L s + Bm J L J t s + ( K1 J L J t + K 2 J L J t + K1 J m J L + K 2 J m J t ) s
5 4 3

+ Bm J L ( K1 + K 2 ) s + K1 K 2 ( J L + J t + J m ) s + Bm K1 K 2 ] = 0
2
(d)Charaacteristicequattion: s ) = 0 .
(
4-446)
(a)Transfferfunction:
Ec ( s ) 1 + R2 Cs
G(s) = =
E (s) 1 + ( R1 + R2 ) Cs
(b)Blockdiagram:
453
(c)Forwardpathtransferfunction:
m (s) K (1 + R2 Cs )
=
E (s) [1 + ( R 1
+ R2 ) Cs ] ( K b K i + Ra J L s )
(d)Closedlooptransferfunction:
m (s) K K (1 + R2 Cs )
=
Fr ( s ) [1 + ( R 1
+ R2 ) Cs ] ( K b K i + Ra J L s ) + K KK e N (1 + R2 Cs )
Ec ( s ) (1 + R Cs )
(e) Gc ( s ) = = 2

E (s) R1Cs
Forwardpathtransferfunction:
m (s) K (1 + R2 Cs )
=
E (s) R1Cs ( K b K i + Ra J L s )
Closedlooptransferfunction:
m (s) K K (1 + R2 Cs )
=
Fr ( s ) R1Cs ( K b K i + Ra J L s ) + K KK e N (1 + R2 Cs )
Ke = 36 pulses / rev = 36 / 2 pulses / rad = 5.73 pulses / rad.

(f) f r = 120 pulses / sec m = 200 RPM = 200(2 / 60 ) rad / sec
f = NK e m = 120 pulses / sec = N ( 36 / 2 ) 200(2 / 60 ) = 120 N pulses / sec
Thus,N=1.For m = 1800 RPM, 120 = N ( 36 / 2 )1800(2 / 60) = 1080 N . Thus, N = 9.
4-47) If the incremental encoder provides a pulse at every edge transition in the two signals of channels
A and B, then the output frequency is increased to four times of input frequency.
4-48)(a)
1 K H (s) K1 H i (s)
1 + K1 H e ( s ) + 1 i H e (s) +
m (s) B + Js Ra + La s B + Js Ra + La s
= =0
TL ( s ) r =0
( s) (s)
Thus,
454
H i (s) H i (s)
H e (s) = = ( Ra + La s )
Ra + La s H e (s)
K1 K i
m ( s) (R + La s )( B + Js )
(b) = a

r (s) TL = 0
( s)
K1 K b K1 H i ( s ) K1 K i K b H e ( s )
( s ) = 1 + K1 H e ( s ) + + +
(R a
+ La s )( B + Js ) Ra + La s (R a
+ La s )( ( B + Js )

K1 K b K1 K i
= 1+ +
(R a
+ La s )( B + Js ) (R a
+ La s )( B + Js )
m (s) K1 K i 1
=
r ( s) TL = 0
(R a
+ La s )( B + Js ) + K i K b + K1 K i K b H e ( s ) Kb H e (s)
4-49)(a)Causeandeffectequations: e = r o e = K s e ea = Ke
dia Ra 1
= ia + (e a
eb ) Tm = K i ia
dt La La
d m Bm d m
2
1 nK L Tm
2
= + Tm ( n m
o ) T2 = 2 = n m
dt J m dt J Jm n
d o
2
KL
2
= ( 2
o )
dt JL
Statevariables: x1 = o , x 2 = o , x3 = m , x 4 = m , x5 = ia
Stateequations:
dx1 dx 2 KL nK L dx3
= x2 = x1 + x3 = x4
dt dt JL JL dt

2
dx 4 nK L n KL Bm Ki dx5 KK s Kb Ra KK s
= x1 x3 x4 + x5 = x1 x4 x5 + r
dt Jm Jm Jm Jm dt La La La La
(b)Statediagram:
455
(c)Forw
wardpathtran
nsferfunction:
o ( s) KK s K i nK L
=
e ( s )
4 3
(
s J m J L La s + J L ( Ra J m + Bm J m + Bm La ) s + n K L La J L + K L J m La + Bm Ra J L s +
2
) 2
(n R K J
2
a L L )
+ Ra K L J m + Bm K L La s + K i K b K L + Ra Bm K L
Clossedlooptranssferfunction:
o ( s) KK s K i nK L
=
r ( s )
5 4
(
J m J L La s + J L ( Ra J m + Bm J m + Bm La ) s + n K L La J L + K L J m La + Bm Ra J L s +
2
) 3
(n R K J
2
a L L )
+ Ra K L J m + Bm K L La s + ( K i K b K L + Ra Bm K L ) s + nKK s K i K L
2
(d) K L = , o = 2 = n m . J L is reflected to motor side so J T = J m + n 2 J L .
Staateequations::
d m Bm Ki d m dia Ra KK s KK s Kb
= m + ia = m = ia + r n m m
dtt JT JT dt dt La La La La
Statediagram:
456
Forrwardpathtraansferfunction
n:
o ( s ) KK s K i n
=
e ( s) s J T La s + ( Ra J T + Bm La ) s + Ra Bm + K i K b
2
Closedlooptranssferfunction:
o ( s) KK s K i n
=
r (s) J T La s + ( Ra J T + Bm La ) s + ( Ra Bm + K i K b ) s + KK s K i n
3 2
Frompaart(c),when K L = , all the terms withoutt K L in o ( s ) / e ( s ) and o ( s ) / r ( s ) cann be neglectedd.
Thesam
meresultsasab
boveareobtained.
4-550)(a)Systemequations:
dv dia dis dis dia

f = K i ia = M T + BT v ea = Ra ia + ( La + Las ) Las + eb 0 = Rs is + ( Ls + Las ) Las
dt dt dt dt dt
(b)TaketheLaplacetraansformonbo
othsidesoftheelastthreeequ
uations,withzeroinitialcond
ditions,wehavve
K i I a ( s ) = ( M T s + BT ) V ( s ) Ea ( s ) = [ Ra + ( La + Las ) s ] I a ( s ) Las sI s ( s ) + K bV ( s )

0 = Las sI a ( s ) + [ Rs + s ( Ls + Las )] I s ( s )
Rearrrangingtheseequations,weeget
Ki V (s) Ki
V (s) = I a (s) Y (s) = = I a (s)
M T s + BT s s ( M T s + BT )

1 Las s
I a (s) = [E ( s ) + Las sI s ( s ) K bV ( s ) ] I s (s) = I a (s)
Ra + ( La + Las ) s Ra + ( La + Las ) s
a
diagram:
Blockd
457
(c)Trannsferfunction::
Y (s) K i [ Rs + ( Ls + Las ) s ]
=
s [ Ra + ( La + Las ) s ] [ Rs + ( Ls + Las ) s ] ( M T s + BT ) + K i K b [ Rs + ( La + Las ) s ] Las s (M s + BT )
2 2
Ea ( s ) T
4-551) (a)Causeeandeffecteqquations:
ea eb
e = r L e = K s e K s = 1 V/rrad ea = Ke ia =
Ra

d m 1 Bm KL d L KL
Tm = K i ia = Tm ( m
L ) = ( m
L ) eb = K b m
dt
d Jm Jm Jm dt JL
15.5
Kb = 15.5 V / KRPM = = 0.148 V / rad / sec
1000 2 / 60
Statteequations:
d L d L KL KL d m d m Bm KL 1 Ki
= L = m L = m = m L + ( KK s e
K b m )
dt dt JL JL dt dt Jm Jm J m Ra
(b)Stattediagram:
(c)Forw
wardpathtran
nsferfunction:
K i KK
Ks KL
G(ss ) =
s J m J L Ra s + ( Bm Ra + K i K b ) J L s + Ra K L ( J L + J m ) s + K L ( Bm Ra + K i K b )
3 2
458
J m Ra J L = 0.03 115
. 0.05 = 0.001725 Bm Ra J L = 10 115
. 0.05 = 0.575 Ki Kb J L = 21 0.148 0.05 = 0.1554
Ra K L J L = 115
. 50000 0.05 = 2875 Ra K L J m = 115
. 50000 0.03 = 1725 Ki KK s K L = 21 1 50000 K = 1050000 K
K L ( Bm Ra + K i K b ) = 50000(10 1.15 + 21 0.148) = 730400
608.7 10 K
6
G(s) =
(
s s + 423.42 s + 2.6667 10 s + 4.2342 10
3 2 6 8
)
(d)Closedlooptransferfunction:
L (s) G(s) K i KK s K L
M ( s ) = = =
r ( s) 1 + G(s) J m J L Ra s + ( Bm Ra + K i K b ) J L s + Ra K L ( J L + J m ) s + K L ( Bm Ra + K i K b ) s + K i KK s K L
4 3 2
8
6.087 10 K
M ( s) = 4 3 6 2 8 8

s + 423.42 s + 2.6667 10 s + 4.2342 10 s + 6.087 10 K
Characteristicequationroots:
K =1 K = 2738 K = 5476
s = 1.45 s = j1000 s = 405 j1223.4

s = 159.88 s = 2117
. j1273.5 s = 617.22 j1275
s = 13105
. j1614.6
4-52)(a)Nonlineardifferentialequations:
dx ( t ) dv ( t )
= v(t ) = k ( v ) g ( x ) + f ( t ) = Bv ( t ) + f ( t )
dt dt
e( t ) e( t 0
With Ra = 0 , ( t ) = = K f i f ( t ) = K f i f ( t ) = K f ia ( t ) Then, ia ( t ) =
Kb v ( t ) Kb K f v ( t )
2
Ki e ( t ) dv ( t ) Ki 2
f ( t ) = Ki ( t )ia ( t ) = 2 2
. Thus, = Bv ( t ) + 2 2
e (t )
Kb K f v (t ) dt Kb K f v (t )
(b)Stateequations: ia ( t ) asinput.
dx ( t ) dv ( t ) 2
= v (t ) = Bv ( t ) + Ki K f ia ( t )
dt dt
(c)Stateequations: ( t ) asinput.
459
2 (t )
f ( t ) = K i K f ia ( t ) ia ( t ) = i f ( t ) =
Kf

dx ( t ) dv ( t ) Ki 2
= v (t ) = Bv ( t ) + (t )
dt dt Kf
4-553)(a)Diffeerentialequations:
d m d m d m d L
2
K i ia = J m + Bm + K ( m L ) + B
dt
2
dt dt
d dt
d m d L = J d L + B d L
2
K ( m L ) + B L 2 + TL
dt dt
L
dt dt
(b)TaketheLaplacetraansformoftheedifferentialeq
quationswithzeroinitialcon
nditions,wegeet
( )
K i I a ( s ) = J m s + Bm s + Bs + K m ( s ) + ( Bs + K ) L ( s )
2

( Bs + K ) m ( )
( s ) ( Bs + K ) L ( s ) = J L s + BL s L ( s ) + TL ( s)
2
s
a L ( s ) fromthelasttwo
Solvingfor m ( s ) and oequations,weehave
Ki Bs + K
m (s) = I a (s) + L (s)
J m s + ( Bm + B ) s + K J m s + ( Bm + B ) s + K
2 2

Bss + K TL ( s )
L (s) = m (s)
J L s + ( BL + B ) s + K J L s + ( BL + B ) s + K
2 2
Signaalflowgraph:
460
(c)Transfermatrix:
m (s) 1 K i J L s + ( BL + B ) s + K Bs + K I a (s)
2
(s) = (s) K i ( Bs + K )

J m s + ( Bm + B ) s + K TL ( s )
L
2
o
o ( s ) = J L J m s + [ J L ( Bm + B ) + J m ( BL + B )] s + [ BL Bm + ( BL + BM ) B + ( J m + J L ) K ] s + K ( BL + B ) s
3 3 2
4-54) As can be estimated by:
1 2
2
1 2
2
Therefore:
2
2 1 1
As a result:
Poles: , ,
zeros:
4-55) By approximating :
1
2
1
2
a)
1
2
1 1
2
Therefore:
1
2
1 1
2
461
b)
1 1
2 2 2 4
1 1
2
3 2
2 2 1 2 1 4 1 2
2 1
4 4
2 1
4-56) MATLAB
clearall
L=1
T=0.1
G1=tf([1/21],conv([0.11],[1/21]))
figure(1)
step(G1)
G2=tf([1144],conv(conv([12],[12]),conv([11],[11])))
figure(2)
step(G2)
L=
1
T=
0.1000
Transferfunction:
0.5s+1
0.05s^2+0.6s+1
Transferfunction:
s^3s^2+4s+4
s^4+6s^3+13s^2+12s+4
462
L
4-57) 423(a)Differentialequations: L ( y ) =
y
d L ( y )i ( t ) dL( y ) dy ( t ) L di ( t ) L dy ( t ) L di ( t )
e ( t ) = Ri ( t ) + = Ri ( t ) + i ( t ) + = Ri ( t ) 2
i(t ) +
dt dy dt y dt y dt y dt
2 2
Ki ( t ) di ( t ) dy ( t ) d y(t )
My ( t ) = Mg 2
At equilibrium, = 0, = 0, 2
= 0
y (t ) dt dt dt
463
Eeq dy eq Eeq K
Thus, ieq = =0 yeq =
R dt R Mg
dy
(b)Definethestatevariablesas x1 = i, x 2 = y, and x3 = .
dt
Eeq Eeq K
Then, x1eq = x 2eq = x3eq = 0
R R Mg
Thedifferentialequationsarewritteninstateequationform:
2
dx1 R x1 x3 x2 dx 2 dx3 K x1
= x1 x 2 + + e = f1 = x3 = f 2 = g 2
= f3
dt L x2 L dt dt M x2
(c)Linearization:
f1 R x3eq Eeq K f1 R x1 x3 Eeq f1 x1eq Mg

= x 2eq + = = x1eq 2
+ =0 = =
x1 L x 2eq L Mg x 2 L x2 L x 3 x 2eq K
f1 x 2 eq 1 K Eeq f 2 f 2 f 2 f 2
= = =0 =0 =1 = 0
e L L Mg R x1 x 2 x 3 e
2
f 3 2 K x1eq 2 Rg f 3 2 K x1eq 2 Rg Mg f 3
= 2
= = 3
= = 0
x1 M x 2 eq Eeq x 2 M x 2eq Eeq K e

Thelinearizedstateequationsabouttheequilibriumpointarewrittenas: x& = A x + B e
Eeq K Mg Eeq K
0
L Mg K
RL Mg
A = 0 B =

0 0 0

2 Rg 2 Rg Mg 0
E 0
eq
Eeq K
4-58)
(a)Differentialequations:
2 2
d y1 ( t ) dy1 ( t ) Ki ( t ) 2 1
M1 2
= M1 g B 2
+ Ki ( t ) 2

dt dt y1 ( t ) y 2 ( t ) y1 ( t )
464
2 2
d y2 (t ) 2 dy 2 ( t ) Ki ( t )
M2 2
= M2 g B 2

dt dt y 2 ( t ) y1 ( t )
dy1 dy 2
Definethestatevariablesas x1 = y1 , x 2 = , x3 = y 2 , x 4 = .
dt dt
Thestateequationsare:
2 2 2
dx1 dx2 Ki Ki dx3 dx4 Ki
= x2 M1 = M 1 g Bx2 + = x4 M2 = M 2 g Bx4
dt dt x1
2
(x 3
x1 )
2
dt dt (x 3
x1 )
2
dx1 dx 2 dx3 dx 4
Atequilibrium, = 0, = 0, = 0, = 0. Thus, x 2eq = 0 and x 4 eq = 0.
dt dt dt dt
2 2 2
KI KI KI
M1g + =0 M2g = 0
X1
2
(X 3
X1 )
2
(X 3
X1 )
2
SolvingforI,with X 1 = 1,wehave
( M1 + M 2 ) g
1/ 2
M + M2
1/ 2
Y2 = X 3 = 1 + 1 I =
M2 K
(b)Nonlinearstateequations:
2 2
dx1 dx2 B K Ki dx3 dx4 B Ki
= x2 =g x2 i + = x4 =g x4
2

M 1 ( x3 x1 ) M 2 ( x3 x1 )
2 2 2
dt dt M1 M 1 x1 dt dt M2
(c)Linearization:
f1 f1 f1 f1 f1
=0 =0 =0 =0 =0
x1 x2 x3 x4 i

f2 f2 f2 2 KI f2
2 2 2
2 KI 2 KI B
= + = = =0
x1 M1 ( X 3 X1 ) x2 x3 M1 ( X 3 X1 ) x4
3 3 3
M 1 x1 M1
f2 2 KI 1 1 f3 f3 f3 f3 f3
= 2+ 2
=0 =0 =0 =1 =0
i M1 X1 ( X 3 X1 ) x1 x2 x3 x4 i
f4 2 KI f4 f4 f4 f4 2 KI
2 2
2 KI B
= =0 = = =
x1 M 2 ( X 3 X1 ) x2 x3 M 2 ( X 3 X1 ) x4 i M 2 ( X 3 X1 )
3 3 2
M2
Linearizedstateequations: M1 = 2, M 2 = 1, g = 32.2, B = 0.1, K = 1.
465
1/ 2
32.2(1 + 2) X = 96.6 X = 9.8285 X
I = X1 =
1
= 1

1 1 1
1 9.8285
( )
X 3 = 1 + 1 + 2 X 1 = 2.732 X 1 = Y2 = 2.732 X 3 X 1 = 1.732
0 1 0 0

2 KI 1 + B 2 KI 0 0
2 2
1 1 0
3
0
M1 X1 ( X 3 X1 )
3
M1 M1 ( X 3 X1 )
3
115.2 0.05 18.59 0
A =

=
0 0 0 1 0 0 0 1

2 KI B 37.18 0 37.18 0.1
2 2
2 KI
0
M 2 ( X 3 X1 ) M 2 ( X 3 X1 ) M 2
3 3
0

2 KI 1 + 1 0
M 1 X 12 ( X 3 X 1 )2 6.552
=

B =
0 0
2 KI 6.552

M 2 ( X 3 X1 )
2
4-59) a)
F3

F1 F2
b) The equation of the translational motion is:
466
The equation of rotational motion is:
where
Also, the relation between rotational and translational motion defines:
Therefore, substituting above expression into the first equation gives:
2
The resulted state space equations are:
2
2
c) According to generalized elements:

1) Viscous friction can be replaced by a resistor where R = B
2) Spring can be replaced by a capacitor where
3) Mass M and m can be replaced by two inductors where and . Then the
angular velocity is measured as a voltage of the inductor L2
4) The gear will be replaced by a transformer with the ratio of
5) The term Mg is also replaced by an input voltage of
467
4-60) As the base is not moving then the model can be reduced to:
Therefore:
1) As , they can be replaced by a inductor with L = m

2) Friction B can be replaced by a resistor where R = B
3) Spring can be replaced by a capacitor where
4) The force F is replaced by a current source where Is = F
4-61)
V2
| |
A
C
g
468
4-62) Recall Eq. (4-324)
Z (s) 1
= 2
&&
Y ( s ) s + 2 n s + n 2
= 1,
Set Y&&( s ) = impulse , pick n for simplicity.
=1
clear all
G=tf([-1],[1 2 1])
figure(1)
impulse(G)
Transfer function:
-1
-------------
s^2 + 2 s + 1
469
4-63)
Use Eq. (4-329).
Kmr
Ra
Z (s) = Va ( s )
La 2 K m Kb
s + 1 ( Js + Bs + K ) + s
Ra Ra
La
s + 1 r
Ra mrY&&( s )
La 2 K m Kb
s + 1 ( Js + Bs + K ) + s
R
a R a
La
For =0 (very small) the format of the equation is similar to Eq. (4-324), and we expect the same
Ra
Kmr
Ra
response for the disturbance input. Except, Z ( s ) = Va ( s ) can be used to reduce the
K K
( Js + Bs + K ) + m b s
2
Ra
effects of disturbance. See Chapter 6.
470
Chapter 5
51(a) 0.707 n 2 rad / sec (b) 0 0.707 n 2 rad / sec
(c) 0.5 1 n 5 rad / sec (d) 0.5 0.707 n 0.5 rad / sec
52(a)Type0 (b)Type0(c)Type1(d)Type2(e)Type3(f)Type3
(g) type 2 (h) type 1
2
53(a) K p = lim G ( s ) = 1000 K v = lim sG ( s ) = 0 K a = lim s G ( s ) = 0
s 0 s 0 s 0
2
(b) K p = lim G ( s ) = K v = lim sG ( s ) = 1 K a = lim s G ( s ) = 0
s 0 s 0 s 0
51
2
(c) K p = lim G ( s ) = K v = lim sG ( s ) = K K a = lim s G ( s ) = 0
s 0 s 0 s 0
2
(d) K p = lim G ( s ) = K v = lim sG ( s ) = K a = lim s G ( s ) = 1
s 0 s 0 s 0
2
(e) K p = lim G ( s ) = K v = lim sG ( s ) = 1 K a = lim s G ( s ) = 0
s 0 s 0 s 0
2
(f) K p = lim G ( s ) = K v = lim sG ( s ) = K a = lim s G ( s ) = K
s 0 s 0 s 0
54(a)Input ErrorConstants SteadystateError
________________________________________________________________________________
u s ( t ) K p = 1000 1 1001
tu s ( t ) K v = 0

2
t us (t ) / 2 K a = 0
(b) Input ErrorConstants SteadystateError
________________________________________________________________________________
u s ( t ) K p = 0
tu s ( t ) K v = 1 1

2
t us (t ) / 2 K a = 0
(c)Input ErrorConstants SteadystateError
________________________________________________________________________________
u s ( t ) K p = 0
52
tu s ( t ) K v = K 1/ K

2
t us (t ) / 2 K a = 0
TheaboveresultsarevalidifthevalueofKcorrespondstoastableclosedloopsystem.
(d)Theclosedloopsystemisunstable.Itismeaninglesstoconductasteadystateerroranalysis.
(e) Input ErrorConstants SteadystateError
________________________________________________________________________________
u s ( t ) K p = 0
tu s ( t ) K v = 1 1

2
t us (t ) / 2 Ka = 0
(f) Input ErrorConstants SteadystateError
________________________________________________________________________________
u s ( t ) Kp = 0
tu s ( t ) K v = 0
2
t us (t ) / 2 K a = K 1/ K
TheclosedloopsystemisstableforallpositivevaluesofK.Thustheaboveresultsarevalid.
G( s) s +1
55(a) K H = H ( 0) = 1 M ( s) = = 3 2

1+ G( s) H ( s) s + 2 s + 3s + 3
a0 = 3, a1 = 3, a2 = 2, b0 = 1, b1 = 1.
UnitstepInput:
1 b0 K H 2
ess = 1 a = 3
KH 0
53
Unitrampinput:
a0 b0 K H = 3 1 = 2 0. Thus e ss = .
UnitparabolicInput:
a0 b0 K H = 2 0 and a1 b1 K H = 1 0. Thus e ss = .
(b) K H = H ( 0) = 5
G( s) 1
M ( s) = = 2
a0 = 5, a1 = 5, b0 = 1, b1 = 0.
1+ G( s) H ( s) s + 5s + 5
UnitstepInput:
1 b0 K H 1 5
ess = 1 a = 5 1 5 = 0
KH 0
UnitrampInput:
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = 5 0
a1 b1 K H 15
e ss = = =
a0 K H 25 5
UnitparabolicInput:
e ss =
(c) K H = H ( 0) = 1 / 5
G( s) s+5
M ( s ) = = 4 3 2
Thesystemisstable.
1+ G( s) H ( s) s + 15s + 50 s + s + 1
a0 = 1, a1 = 1, a2 = 50, a3 = 15, b0 = 5, b1 = 1
UnitstepInput:
1 b0 K H 5/5
ess = 1 a = 5 1 1 = 0
KH 0
UnitrampInput:
54
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = 4 / 5 0
a1 b1 K H 1 1/ 5
e ss = = = 4
a0 K H 1/ 5
UnitparabolicInput:
e ss =
(d) K H = H ( 0 ) = 10
G( s) 1
M ( s ) = = 3 2
Thesystemisstable.
1+ G( s) H ( s) s + 12 s + 5s + 10
a0 = 10, a1 = 5, a2 = 12, b0 = 1, b1 = 0, b2 = 0
UnitstepInput:
1 b0 K H 1 10
ess = 1 a = 10 1 10 = 0
KH 0
UnitrampInput:
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = 5 0
a1 b1 K H 5
e ss = = = 0.05
a0 K H 100
UnitparabolicInput:
e ss =
s+4
56(a) M ( s ) = 4 3 2
K H = 1 Thesystemisstable.
s + 16 s + 48 s + 4 s + 4
a0 = 4, a1 = 4, a2 = 48, a3 = 16, b0 = 4, b1 = 1, b2 = 0, b3 = 0
UnitstepInput:
1 b0 K H 4
ess = 1 a = 1 4 = 0
KH 0
Unitrampinput:
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = 4 1 = 3 0
55
a1 b1 K H 4 1 3
e ss = = =
a0 K H 4 4
UnitparabolicInput:
e ss =
K ( s + 3)
(b) M ( s ) = 3 2
K H = 1 Thesystemisstablefor K > 0.
s + 3s + ( K + 2) s + 3K
a0 = 3 K , a1 = K + 2, a2 = 3, b0 = 3 K , b1 = K
UnitstepInput:
1 b0 K H 3K
ess = 1 a = 1 3K = 0
KH 0
UnitrampInput:
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = K + 2 K = 2 0
a1 b1 K H K + 2 K 2
e ss = = =
a0 K H 3K 3K
UnitparabolicInput:
e ss =
TheaboveresultsarevalidforK>0.
s+5 10 s H ( s)
(c) M ( s ) = 4 3 2
H ( s) = K H = lim = 2
s + 15 s + 50 s + 10 s s+5 s 0 s
a0 = 0, a1 = 10, a2 = 50, a3 = 15, b0 = 5, b1 = 1
UnitstepInput:
1 a2 b1 K H 1 50 1 2
ess = a = = 2.4
KH 1 2 10
UnitrampInput:
56
e ss =
UnitparabolicInput:
e ss =
K ( s + 5)
(d) M ( s ) = 4 3 2
K H = 1 Thesystemisstablefor0<K<204.
s + 17 s + 60 s + 5 Ks + 5 K
a0 = 5 K , a1 = 5K , a2 = 60, a3 = 17, b0 = 5K , b1 = K
UnitstepInput:
b0 K H
1 5K
ess = 1 a = 1 5K = 0
KH 0
UnitrampInput:
i = 0: a0 b0 K H = 0 i = 1: a1 b1 K H = 5 K K = 4 K 0
a1 b1 K H 5K K 4
e ss = = =
a0 K H 5K 5
UnitparabolicInput:
e ss =
Theresultsarevalidfor0<K<204.
57
5-7)
1
3 2 5 1
5 1 5 11 5
1
2 3
Type of the system is zero
Pole: s = -2.2013 + 1.8773i, s = -2.2013 - 1.8773i, and s = -0.5974
Zero: s = -1
5-8)
5( s + 1)
s ( s + 2)( s + 3)
5 1
2 3
a) Position error: lim G s lim
b) Velocity error: lim lim
c) Acceleration error: lim lim 0
5-9) a) Steady state error for unit step input:
Referring to the result of problem 5-8, 0
b) Steady state error for ramp input:
58
Regarding the result of problem 5-8,
c) Steady state error for parabolic input:
Regarding the result of problem 5-8, 0
5-10)
4( s + 1)
s 2 ( s + 2)
a) Step error constant: lim
b) Ramp error constant: lim
c) Parabolic error constant: lim 2
5-11) 3 4
where x1 is a unit step input, x2 is a ramp input, and x3 is a unit parabola input. Since the system
is linear, then the effect of X(s) is the summation of effect of each individual input.
That is: 3 4
So:
4 2
59
5-12) The step input response of the system is:
1 1 1 1
1
Therefore:
1
1
1
The rise time is the time that unit step response value reaches from 0.1 to 0.9. Then:
1 . .
1
It is obvious that tr > 0, then:
1 . .
0
1
As |k| < 1, then 0
. . . .
Therefore 0 or
which yields: 0
513)
Y ( s) KG p ( s ) 20 s 100 K
G( s) = = = Type1system.
E ( s) 1 + Kt G p ( s ) 20 s(1 + 0.2 s + 100 Kt )
5K
Errorconstants: K p = , K v = , Ka = 0
1 + 100 Kt
1
(a) r ( t ) = u s ( t ): e ss = = 0
1+ K p
1 1 + 100 Kt
(b) r ( t ) = tu s ( t ): e ss = =
Kv 5K
1
(c) r ( t ) = t 2u s ( t ) / 2:e ss = =
Ka
510
514
100 Y ( s) KG p ( s )
G p ( s) = G( s) = =
(1 + 0.1s )(1 + 0.5s ) E ( s) 20 s 1 + Kt G p ( s )
100 K
G( s) =
20 s (1 + 0.1s )(1 + 0.5 s ) + 100 Kt
5K
Errorconstants: K p = , K v = , Ka = 0
1 + 100 Kt
1
(a) r ( t ) = u s ( t ): e ss = = 0
1+ K p
1 1 + 100 Kt
(b) r ( t ) = tu s ( t ): e ss = =
Kv 5K
1
(c) r ( t ) = t 2u s ( t ) / 2:e ss = =
Ka
Sincethesystemisofthethirdorder,thevaluesofKand Kt mustbeconstrainedsothatthesystemis
stable.Thecharacteristicequationis
s + 12 s + ( 20 + 2000 K t ) s + 100 K = 0
3 2

RouthTabulation:
3
s 1 20 + 2000 Kt
2
s 12 100 K
240 + 24000 Kt 100 K
1
s
12
0
s 100 K
1 + 100 K t 1
StabilityConditions: K>0 12 (1 + 100 K t ) 5 K > 0 or >
5K 12
Thus,theminimumsteadystateerrorthatcanbeobtainedwithaunitrampinputis1/12.
511
515(a)FromFigure3P29,
K1 K 2 K i K b + KK1 K i K t
1+ +
o ( s) Ra + La s (R + La s )( Bt + J t s )
= a

r (s) K1 K 2 K i K b + KK1 K i K t KK s K1 K i N
1+ + +
Ra + La s (Ra
+ La s )( Bt + J t s ) s ( Ra + La s )( Bt + J t s )
o (s) s [( Ra + La s )( Bt + J t s ) + K1 K 2 ( Bt + J t s ) + K i K b + KK1 K i K t ]
=
r (s) La J t s + ( La Bt + Ra J t + K1 K 2 J t ) s + ( Ra Bt + K i K b + KK i K1 K t + K1 K 2 Bt ) s + KK s K1 K i N
3 2
1
r ( t ) = u s ( t ), r ( s ) = lim s e ( s ) = 0
s s 0
Providedthatallthepolesof s e ( s ) areallinthelefthalfsplane.
2
(b)Foraunitrampinput, r ( s ) = 1 / s .
Ra Bt + K1 K 2 Bt + Ki Kb + KK1 Ki Kt
e ss = lim e ( t ) = lim s e ( s ) =
t s 0 KK s K1 Ki N
ifthelimitisvalid.
512
516(a)Forwardpathtransferfunction:[n(t)=0]:
K (1 + 0.02 s )
Y ( s) s ( s + 25)
2
K (1 + 0.02 s )
G (s) = = = Type1system.
E (s)
1+ 2
KK t s (
s s + 25s + KK t
2
)
s ( s + 25)
1
ErrorConstants: K p = , Kv = , Ka = 0
Kt
1 1
Foraunitrampinput, r ( t ) = tu s ( t ), R( s ) = 2
, e ss = lim e( t ) = lim sE ( s ) = = Kt
s t s 0 Kv
RouthTabulation:
KK t + 0.02 K
3
s 1
2
s 25 K

1 25 K ( K t + 0.02 ) K
s
25
0
s K
StabilityConditions: K >0 25 ( K t + 0.02 ) K > 0 or K t > 0.02
(b)Withr(t)=0, n( t ) = u s ( t ), N ( s ) = 1 / s.
SystemTransferFunctionwithN(s)asInput:
K
Y (s) s ( s + 25) K
2
= = 3
K (1 + 0.02 s ) KK t s s + 25 s + K ( K t + 0.02 ) s + K
2
N (s)
1+ 2 + 2
s ( s + 25) s ( s + 25)
SteadyStateOutputdueton(t):
y ss = lim y ( t ) = lim sY ( s ) = 1 ifthelimitisvalid.

t s 0
513
517YoumayuseMATLABinallRouthHurwitzcalculations.
1. ActivateMATLAB
3. Typein
Acsys
4. ThenpressthetransferfunctionSymbolicandentertheCharacteristicequation
5. ThenpresstheRouthHurwitzbutton
6. ForexamplelookatbelowFigures
514
(a) n( t ) = 0, r ( t ) = tu s ( t ).
ForwardpathTransferfunction:
Y (s) K ( s + )( s + 3)
G(s) = = Type1system.
E ( s) n=0 (
s s 1
2
)
Ramperrorconstant: K v = lim sG ( s ) = 3 K
s 0
1 1
Steadystateerror: e ss = =
Kv 3Kv
3 2
Characteristicequation: s + Ks + [ K ( 3 + ) 1] s + 3K = 0
RouthTabulation:
3
s 1 3 K + K 1
2
s K 3K
K ( 3 K + K 1) 3K
1
s
K
0
s 3K
1 + 3K
3 K + K 1 3 > 0 or K>
StabilityConditions: 3+
K > 0
(b)Whenr(t)=0, n( t ) = u s ( t ), N ( s ) = 1 / s.
K ( s + 3)
Y (s) s 1
2
Ks ( s + 3)
TransferFunctionbetweenn(t)andy(t): = =
N (s) K ( s + )( s + 3) s 3 + Ks 2 + [ K ( s + ) 1]s + 3 K
r =0
1+
s s 1
2
( )
SteadyStateOutputdueton(t):
y ss = lim y ( t ) = lim sY ( s ) = 0 ifthelimitisvalid.

t s 0
515
518

2
1
Percent maximum overshoot = 0.25 = e
Thus
1 = ln 0.25 = 1.386 = 1.922 (1 )

2 2 2 2

Solvingfor fromthelastequation,wehave =0.404.

Peak Time t max = = 0.01 sec.Thus, n = = 343.4 rad / sec
2 2
n 1 0.01 1 ( 0.404)
TransferFunctionoftheSecondorderPrototypeSystem:
2
Y ( s) n 117916
= 2 2
= 2

R( s ) s + 2 n s + n s + 277.3s + 117916
516
ExtendedMATLABsolutionsofproblemssimilarto519527appearlater
one.g.558
519ClosedLoopTransferFunction: Characteristicequation:
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
Forasecondorderprototypesystem,whenthemaximumovershootis4.3%, = 0.707.
n = 25 K , 2 n = 5 + 500 Kt = 1.414 25 K
RiseTime:
2
1 0.4167 + 2.917 2.164
tr = = = 0.2 sec Thus n = 10.82 rad / sec
n n
2
n (10.82)
2
10.3
Thus, K= = = 4.68 5 + 500 Kt = 1.414 n = 15.3 Thus Kt = = 0.0206
25 25 500
With K = 4.68 and Kt = 0.0206, the system transfer function is
Y ( s) 117
= 2

R( s ) s + 15.3s + 117
UnitstepResponse:
y=0.1att=0.047sec.
y=0.9att=0.244sec.
t r = 0.244 0.047 = 0.197 sec.
y max = 0.0432 ( 4.32% max. overshoot)
517
520ClosedloopTransferFunction: CharacteristicEquation:
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)

= ln 0.1 = 2.3 = 5.3 (1 )
2 2 2
WhenMaximumovershoot=10%,
1
2
Solvingfor ,weget =0.59.
TheNaturalundampedfrequencyis n = 25K Thus, 5+ 500 Kt = 2 n = 118

. n
RiseTime:
2
1 0.4167 + 2.917 17696
.
tr = = 0.1 = sec. Thus n = 17.7 rad / sec
n n
2
n 15.88
K= = 12.58 Thus Kt = = 0.0318
25 500
WithK=12.58and Kt = 0.0318, thesystemtransferfunctionis
Y ( s) 313
= 2

R( s ) s + 20.88 s + 314.5
UnitstepResponse:
y=0.1whent=0.028sec.
t r = 0.131 0.028 = 0.103 sec.
y max = 11
. (10% max. overshoot )
518
521 ClosedLoopTransferFunction: CharacteristicEquation:
Y ( s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2
2

R( s ) s + ( 5 + 500 Kt ) s + 25 K

= ln 0.2 = 1.61 = 2.59 (1 )
2 2 2
WhenMaximumovershoot=20%,
1
2
Solvingfor ,weget = 0.456.
TheNaturalundampedfrequency n = 25 K 5 + 500 Kt = 2 n = 0.912 n
RiseTime:
2
1 0.4167 + 2.917 1.4165 1.4165
tr = = 0.05 = sec. Thus, n = = 28.33
n n 0.05
2
n
K= = 32.1 5 + 500 Kt = 0.912 n = 25.84 Thus, Kt = 0.0417
25
WithK=32.1and Kt = 0.0417, thesystemtransferfunctionis
Y ( s) 802.59
= 2

R( s ) s + 25.84 s + 802.59
UnitstepResponse:
t r = 0.072 0.0178 = 0.0542 sec.
y max = 1.2 ( 20% max. overshoot )
519
522ClosedLoopTransferFunction: CharacteristicEquation:
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
2
. + 0.125 + 0.469
11
Delaytime t d = 0.1 sec.
n
1.423
WhenMaximumovershoot=4.3%, = 0.707. td = = 0.1 sec. Thus n = 14.23rad/sec.
n
2 2
K =
n = 14.23 = 8.1 5 + 500 K = 2 = 1.414 = 20.12 Thus K = 1512
.
= 0.0302

5 5 t
t n n
500
WithK=20.12and Kt = 0.0302 ,thesystemtransferfunctionis
Y ( s) 202.5
= 2

R( s ) s + 20.1s + 202.5
UnitStepResponse:
Wheny=0.5,t=0.1005sec.
Thus, t d = 0.1005 sec.
y max = 1043
. ( 4.3% max. overshoot )
520
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
2
. + 0.125 + 0.469
11 1337
. 1337
.
Delaytime t d = 0.05 = Thus, n = = 26.74
n n 0.05
2 2
26.74 = 28.6 5 + 500 K = 2 = 2 0.59 26.74 = 3155
K = n = Thus Kt = 0.0531
.
5 5 t n
WithK=28.6and Kt = 0.0531, the system transfer function is
Y ( s) 715
= 2

R( s ) s + 3155
. s + 715
UnitStepResponse:
Thus, t d = 0.0505sec.
y max = 11007
. (10.07% max. overshoot )
521
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
ForMaximumovershoot=0.2, = 0.456 .
2
. + 0.125 + 0.469
11 1.2545
Delaytime t d = = = 0.01sec.
n n
NaturalUndampedFrequency n =
1.2545 15737.7 = 629.5
= 125.45rad/sec.Thus, K = n =
0.01 5 25
5 + 500 Kt = 2 n = 2 0.456 125.45 = 114.41 Thus, Kt = 0.2188
WithK=629.5and Kt = 0.2188 ,thesystemtransferfunctionis
Y ( s) 15737.7
= 2

R( s ) s + 114.41s + 15737.7
UnitstepResponse:
Thus, t d = 0.0101sec.
y max = 1.2 ( 20% max. overshoot )
522
Y (s) 25 K
= s + ( 5 + 5000 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
= 0.6 2 n = 5 + 500 Kt = 1.2 n
3.2 3.2 3.2

Settlingtime t s = = 0.1sec.Thus, n = = 53.33 rad / sec
n 0.6 n 0.06
2
1.2 n 5 n
Kt = = 0.118 K= = 113.76
500 25
SystemTransferFunction:
Y ( s) 2844
= 2

R( s ) s + 64 s + 2844
UnitstepResponse:
y(t)reaches1.00andneverexceedsthis
valueatt=0.098sec.
Thus, t s = 0.098 sec.
523
526(a)ClosedLoopTransferFunction: CharacteristicEquation:
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
Formaximumovershoot=0.1, = 0.59. 5 + 500 Kt = 2 n = 2 0.59 n = 118

. n
3.2 3.2 3.2

Settlingtime: t s = = = 0.05 sec. n = = 108.47
n 0.59 n 0.05 0.59
2
. n 5
118 n
Kt = = 0.246 K= = 470.63
500 25
Y ( s) 11765.74
= 2

R( s ) s + 128 s + 11765.74
UnitStepResponse:
y(t)reaches1.05andneverexceeds
thisvalueatt=0.048sec.
Thus, t s = 0.048 sec.
(b)Formaximumovershoot=0.2, = 0.456. 5 + 500 Kt = 2 n = 0.912 n
3.2 3.2 3.2

Settlingtime t s = = = 0.01 sec. n = = 701.75 rad / sec
n 0.456 n 0.456 0.01
0.912 n 5
Kt = = 1.27
500
Y ( s) 492453
= 2

R( s ) s + 640 s + 492453
UnitStepResponse:
524
y(t)reaches1.05andnever
exceedsthisvalueatt=0.0074sec.
Thus, t s = 0.0074 sec. Thisisless
thanthecalculatedvalueof0.01sec.
525
Y (s) 25 K
= s + ( 5 + 500 K t ) s + 25 K = 0
2

s + ( 5 + 500 K t ) s + 25 K
2
R(s)
4.5 3.1815
Dampingratio = 0.707. Settlingtime t s = = = 0.1sec. Thus, n = 31815
. rad/sec.
n n
2
n
5 + 500 Kt = 2 n = 44.986 Thus, Kt = 0.08 K= = 40.488
2
Y ( s) 1012.2
= 2

R( s ) s + 44.986 s + 1012.2
UnitStepResponse:Theunitstepresponsereaches0.95att=0.092sec.whichisthemeasured t s .
526
528(a)When = 0.5 ,therisetimeis

2
1 0.4167 + 2.917 1521
.
tr = = 1 sec.Thus n = 1.521rad/sec.
n n
Thesecondordertermofthecharacteristicequationiswritten
2 2 2
s + 2 n s + n = s + 1521
. s + 2.313 = 0
3 2
Thecharacteristicequationofthesystemis s + ( a + 30 ) s + 30 as + K = 0
2
Dividingthecharacteristicequationby s + 1521
. s + 2.313, wehave
Forzeroremainders, 28.48 a = 45.63 Thus, a = 16

. K = 65.874 + 2.313a = 69.58
ForwardPathTransferFunction:
69.58
G( s) =
s( s + 16
. )( s + 30)
UnitStepResponse:
y=0.9whent=1.43sec.
RiseTime:
t r = 1.43 0.355 = 1075

. sec.
527
(b)Thesystemistype1.
(i)Foraunitstepinput, e ss = 0.
K 60.58 1
(ii)Foraunitrampinput, K v = lim sG ( s ) = = = 1.45 e ss = = 0.69
s 0 30 a 30 16
. Kv
529(a)CharacteristicEquation:
3 2
s + 3 s + ( 2 + K ) s K = 0
ApplytheRouthHurwitzcriteriontofindtherangeofKforstability.
RouthTabulation:
3
s 1 2+ K
2
s 3 K
6 + 4K
1
s
3
0
s K
StabilityCondition: 1.5<K<0ThissimplifiesthesearchforKfortwoequalroots.
WhenK=0.27806,thecharacteristicequationrootsare:0.347,
0.347,and2.3054.
(b)UnitStepResponse:(K=0.27806)
528
(c)UnitStepResponse(K=1)

Thestepresponsesin(a)and(b)allhaveanegativeundershootforsmallvaluesoft.Thisisduetothe
zeroofG(s)thatliesintherighthalfsplane.
529
530(a)Thestateequationsoftheclosedloopsystemare:
dx1 dx 2
= x1 + 5x 2 = 6 x1 k1 x1 k 2 x 2 + r
dt dt
Thecharacteristicequationoftheclosedloopsystemis
s +1 5
= = s + (1 + k 2 ) s + ( 30 + 5k1 + k 2 ) = 0
2

6 + k1 s + k2
2
For n = 10 rad / sec, 30 + 5k1 + k 2 = n = 100. Thus 5k1 + k 2 = 70
k2
(b)For = 0.707, 2 n = 1 + k 2 . Thus n = 1 + .
1.414
(1 + k ) 2
n = = 30 + 5k1 + k 2 k 2 = 59 + 10 k1
2 2 2
Thus
2
(c)For n = 10 rad / sec and = 0.707,
5k1 + k 2 = 100 and 1 + k 2 = 2 n = 14.14 Thus k 2 = 1314

.
Solvingfor k1 , we have k1 = 1137

. .
(d)Theclosedlooptransferfunctionis
Y (s) 5 5
= =
s + ( k 2 + 1) s + ( 30 + 5k1 + k 2 ) s + 14.14 s + 100
2 2
R(s)
5
Foraunitstepinput, lim y ( t ) = lim sY ( s ) = = 0.05
t s 0 100
(e)Forzerosteadystateerrorduetoaunitstepinput,
30 + 5k1 + k 2 = 5 Thus 5k1 + k 2 = 25
530
ParameterPlane k1versus k 2 :
531(a)ClosedLoopTransferFunction (b)CharacteristicEquation:
Y (s) 100 ( K P + K D s ) 2
= s + 100 K D s + 100 K P = 0
s + 100 K D s + 100 K P
2
R(s)
Thesystemisstablefor K P > 0 and K D > 0.
(b)For = 1, 2 n = 100 K D .
n = 10 K P Thus 2 n = 100 K D = 20 K P K D = 0.2 K P
(c)Seeparameterplaneinpart(g).
531
(d)Seeparameterplaneinpart(g).
2
(e)Parabolicerrorconstant K a = 1000 sec
K a = lim s G ( s ) = lim100 ( K P + K D s ) = 100 K P = 1000 Thus K P = 10

2

s0 s 0
(f)Naturalundampedfrequency n = 50 rad/sec.
n = 10 K P = 50 Thus K P = 25
(g)When K P = 0,
100 K D s 100 K D
G( s) = 2
= (polezerocancellation)
s s
532
532(a)ForwardpathTransferFunction:
Y (s) KK i 10 K
G ( s ) = = =
E (s) s [ Js (1 + Ts ) + K i K t ] (
s 0.001s + 0.01s + 10 K t
2
)
K 1 Kt
When r ( t ) = tu s ( t ), K v = lim sG ( s ) = e ss = =
s 0 Kt Kv K
(b)Whenr(t)=0
Y (s) 1 + Ts 1 + 0.1s
= =
Td ( s ) s [ Js (1 + Ts ) + K i K t ] + KK i (
s 0.001s + 0.01s + 10 K t + 10 K
2
)
1 1
For Td ( s ) = lim y ( t ) = lim sY ( s ) = ifthesystemisstable.
s t s 0 10 K
(c)Thecharacteristicequationoftheclosedloopsystemis
3 2
0.001s + 0.01s + 0.1s + 10 K = 0
ThesystemisunstableforK>0.1.SowecansetKtojustlessthan0.1.Then,theminimumvalueof
thesteadystatevalueofy(t)is
1 +
=1
10 K K = 0.1

However,withthisvalueofK,thesystemresponsewillbeveryoscillatory.Themaximumovershoot
willbenearly100%.
(d)ForK=0.1,thecharacteristicequationis
3 2 3 2 4
0.001s + 0.01s + 10 Kt s + 1 = 0 or s + 10 s + 10 Kt s + 1000 = 0
533
Forthetwocomplexrootstohaverealpartsof2/5.weletthecharacteristicequationbewrittenas
(
( s + a ) s + 5s + b = 0
2
) or s + ( s + 5) s + (5a + b) s + ab = 0
3 2
4
Then, a + 5 = 10 a=5 ab = 1000 b = 200 5a + b = 10 Kt Kt = 0.0225
Thethreerootsare: s = a = 5 s = a = 5 s = 2.5 j13.92
. . . . .
5-33) Rise time: 0.56
.
Peak time: 0.785
.
.
Maximum overshoot: . 0.095
.
Settling time: 0 0.69
.
1.067
.
534
5-34)
K
s ( Js + a + KK f )
exp 0.2 = 0.456

0.1 n = 0.353
0.125

2 5.42
. .
5.49
.
19.88
535
5-35) a)
6.5
1
6.5 2
.
1
6.5 1

. .
Substituting into equation (2) gives:

7.5 6.5
6.5 6.5
Since the system is multi input and multi output, there are 4 transfer functions as:
, , ,
To find the unit step response of the system, lets consider

6.5
6.5 2
6.5 6.5
1 1
2 1
2 26.5
By looking at the Laplace transform function table:
1
1 sin 1
1
where cos
b)
536
Therefore:
As a result:
which means:
1
1
The unit step response is:
1
1
Therefore as a result:
1
1 sin 1
1
where n = 1 and n = 1 / 2
c)
Therefore:
1 1
As a result, the step response of the system is:
1
1
By looking up at the Laplace transfer function table:
537
where n = 1, and 2 n = 1 = 1 / 2
sin 1
where = cos-1 (2 2 -1) = cos-1(0.5), therefore,
2 3
1 sin
3 2
5-36) MATLAB CODE

(a)
clear all
Amat=[-1 -1;6.5 0]
Bmat=[1 1;1 0]
Cmat=[1 0;0 1]
Dmat=[0 0;0 0]
disp(' State-Space Model is:')
Statemodel=ss(Amat,Bmat,Cmat,Dmat)
[mA,nA]=size(Amat);
rankA=rank(Amat);
disp(' Characteristic Polynomial:')
chareq=poly(Amat);
%p = poly(A) where A is an n-by-n matrix returns an n+1 element

%row vector whose elements are the coefficients of the characteristic
%polynomialdet(sI-A). The coefficients are ordered in descending powers.
[mchareq,nchareq]=size(chareq);
syms 's';
poly2sym(chareq,s)
disp(' Equivalent Transfer Function Model is:')
Hmat=Cmat*inv(s*eye(2)-Amat)*Bmat+Dmat
Since the system is multi input and multi output, there are 4 transfer functions as:
, , ,
To find the unit step response of the system, lets consider
6.5
6.5
538
Lets obtain this term and find Y2(s) time response for a step input.
H22=Hmat(2,2)
ilaplace(H22/s)
Pretty(H22)
H22poly=tf([13/2],chareq)
step(H22poly)
H22 =
13/(2*s^2+2*s+13)
ans =
1-1/5*exp(-1/2*t)*(5*cos(5/2*t)+sin(5/2*t))
13
---------------
2
2 s + 2 s + 13
Transfer function:
6.5
-------------
s^2 + s + 6.5
To find the step response H11, H12, and H21 follow the same procedure.
Other parts are the same.
539
5-37) Impulse response:

.
a) and 1 , therefore,
.
sin 1
1
b)
sin 1
1
c)
1
1 sin 1
1
where = cos-1
5-38) Use the approach in 5-36 except:

H22=Hmat(2,2)
ilaplace(H22)
Pretty(H22)
H22poly=tf([13/2],chareq)
impulse(H22poly)
H22 =
13/(2*s^2+2*s+13)
ans =
13/5*exp(-1/2*t)*sin(5/2*t)
13
---------------
2
2 s + 2 s + 13
Transfer function:
6.5
-------------
s^2 + s + 6.5
540
Other parts are the same.
5-39) a) The displacement of the bar is:
sin
Then the equation of motion is:
As x is a function of and changing with time, then
cos
If is small enough, then sin and cos 1. Therefore, the equation of motion is
rewritten as:
541
(b) To find the unit step response, you can use the symbolic approach shown in Toolbox 2-1-1:
clear all
%s=tf(s);
syms s B L K
Theta=B*s/s/L/(B*s+K)
ilaplace(Theta)
Theta =
B/L/(B*s+K)
ans =
1/L*exp(-K*t/B)
Alternatively, assign values to B L K and find the step response see solution to problem 5-36.
540(a) Kt = 10000 oz - in / rad
TheForwardPathTransferFunction:
9 10 K
12
G(s) =

(
s s + 5000 s + 1.067 10 s + 50.5 10 s + 5.724 10
4 3 7 2 9 12
)
9 10 K
12
=
s ( s + 116)( s + 4883)( s + 41.68 + j 3178.3)( s + 41.68 j 3178.3)
RouthTabulation:
5 7 12
s 1 1.067 10 5.724 10
4 9 12
s 5000 50.5 10 9 10 K

3 5 12 9
s 5.7 10 5.72 10 18
. 10 K 0
2 8 7 12
s 2.895 10 + 1579
. 10 K 9 10 K
13 12 9 2
1 16.6 10 + 8.473 10 K 2.8422 10 K
s
29 + 1579
. K
0 12
s 9 10 K
542
1 2
Fromthe s row,theconditionofstabilityis165710 + 8473 K 2.8422 K > 0
2
or K 298114
. K 58303.427 < 0 or ( K + 19.43)( K 3000.57 ) < 0
StabilityCondition: 0<K<3000.56
ThecriticalvalueofKforstabilityis3000.56.WiththisvalueofK,therootsofthecharacteristic
equationare: 4916.9,41.57+j3113.3,41.57+j3113.3,j752.68,andj752.68
(b) K L = 1000 ozin/rad.Theforwardpathtransferfunctionis
9 10 K
11
G(s) =

(
s s + 5000 s + 1.582 10 s + 5.05 10 s + 5.724 10
4 3 6 2 9 11
)
9 10 K
11
=
s (1 + 116.06)( s + 4882.8)( s + 56.248 + j1005)( s + 56.248 j1005)
(c)CharacteristicEquationoftheClosedLoopSystem:
5 4 6 3 9 2 11 11
s + 5000 s + 1.582 10 s + 5.05 10 s + 5.724 10 s + 9 10 K = 0
RouthTabulation:
5 6 11
s 1 1.582 10 5.724 10
4 9 11
s 5000 5.05 10 9 10 K

3 5 11 8
s 5.72 10 5.724 10 18
. 10 K 0
2 7 6 11
s 4.6503 10 + 15734
. 10 K 9 10 K
18 15 14 2
1 26.618 10 + 377.43 10 K 2.832 10 K
s 7 6
4.6503 10 + 15734
. 10 K
0 11
s 9 10 K
543
1 4 2
Fromthe s row,theconditionofstabilityis 26.618 10 + 3774.3 K 2.832 K > 0
2
Or, K 1332.73 K 93990 < 0 or ( K 1400 )( K + 67.14 ) < 0
StabilityCondition: 0<K<1400
ThecriticalvalueofKforstabilityis1400.WiththisvalueofK,thecharacteristicequationrootare:
4885.1,57.465+j676,57.465j676,j748.44,andj748.44
(c) K L = .
ForwardPathTransferFunction:
nK s K i K
G(s) = JT = J m + n J L
2

s La J T s + ( Ra J T + Rm La ) s + Ra Bm + K i K b
2
891100 K 891100 K
= =
(
s s + 5000 s + 566700
2
) s ( s + 116)( s + 4884)
CharacteristicEquationoftheClosedLoopSystem:
3 2
s + 5000 s + 566700 s + 891100 K = 0
RouthTabulation:
3
s 1 566700
2
s 5000 891100 K

1
s 566700 178.22 K
0
s 8991100 K
1
Fromthe s row,theconditionofKforstabilityis566700178.22K>0.
544
StabilityCondition: 0<K<3179.78
ThecriticalvalueofKforstabilityis3179.78.WithK=3179.78,thecharacteristicequationrootsare
5000,j752.79,andj752.79.
Whenthemotorshaftisflexible, K L isfinite,twooftheopenlooppolesarecomplex.Astheshaft
becomesstiffer, K L increases,andtheimaginarypartsoftheopenlooppolesalsoincrease.When
K L = ,theshaftisrigid,thepolesoftheforwardpathtransferfunctionareallreal.Similareffects
areobservedfortherootsofthecharacteristicequationwithrespecttothevalueof K L .
541(a)
100( s + 2 )
Gc ( s ) = 1 G( s) = 2
K p = lim G ( s ) = 200
s 1 s 0
Whend(t)=0,thesteadystateerrorduetoaunitstepinputis
1 1 1
e ss = = = = 0.005025
1+ K p 1 200 199
(b)
s + 100( s + 2)( s + )
Gc ( s ) = G(s) = Kp = ess = 0
s (
s s 1
2
)
(c)
=5 maximum overshoot = 5.6%
= 50 maximum overshoot = 22%
= 500 maximum overshoot = 54.6%
545
Asthevalueof increases,themaximumovershootincreasesbecausethedampingeffectofthe
zeroat s = becomeslesseffective.
UnitStepResponses:
1
(d) r ( t ) = 0 and Gc ( s ) = 1. d ( t ) = u s ( t ) D( s ) =
s
SystemTransferFunction:(r=0)
Y ( s) 100( s + 2)
= 3 2

D( s ) r =0
s + 100 s + (199 + 100 ) s + 200
OutputDuetoUnitStepInput:
100( s + 2 )
Y ( s) =
3 2
s s + 100 s + (199 + 100 ) s + 200
200 1
y ss = lim y ( t ) = lim sY ( s ) = =
t s 0 200
(e) r ( t ) = 0, d ( t ) = u s ( t )
s +
Gc ( s ) =
s
546
SystemTransferFunction[r(t)=0]
Y ( s) 100 s( s + 20 1
= 3 2
D( s ) =
D( s ) r =0
s + 100 s + (199 + 100 ) s + 200 s
y ss = lim y ( t ) = lim sY ( s ) = 0
t s 0
(f)
Y ( s) 100 s( s + 2 )
=5 = 3 2
D( s ) r =0
s + 100 s + 699 s + 1000
Y ( s) 100 s( s + 2)
= 50 = 3 2

D( s ) r =0
s + 100 s + 5199 s + 10000
Y ( s) 100 s( s + 2)
= 5000 = 3 2
D( s ) r =0
s + 100 s + 50199 s + 100000
UnitStepResponses:
(g)Asthevalueof increases,theoutputresponsey(t)duetor(t)becomesmoreoscillatory,andthe
overshootislarger.Asthevalueof increases,theamplitudeoftheoutputresponsey(t)duetod(t)
becomessmallerandmoreoscillatory.
547
542(a)ForwardPathTransferfunction: CharacteristicEquation:
H ( s) 10 N N 2
G ( s ) = = s + s + N = 0
E ( s) s( s + 1)( s + 10) s( s + 1)
2
N=1:CharacteristicEquation: s + s + 1 = 0 = 0.5 n = 1rad/sec.

1
2

Maximum overshoot = e = 0.163 (16.3%)Peaktime t max = = 3.628 sec.
2
n 1
2
N=10:CharacteristicEquation: s + s = 10 = 0 = 0.158 n = 10 rad/sec.

1
2

Maximumovershoot= e = 0.605 (60.5%)Peaktime t max = = 1006
. sec.
2
n 1
(b)UnitStepResponse:N=1
548
SecondorderSystem ThirdorderSystem
Maximumovershoot 0.163 0.206
Peaktime 3.628sec. 3.628sec.
UnitStepResponse:N=10
SecondorderSystem ThirdorderSystem
Maximumovershoot 0.605 0.926
Peaktime 1.006sec. 1.13sec.
549
543UnitStepResponses:
When Tz issmall,theeffectislowerovershootduetoimproveddamping.WhenTz isverylarge,the
overshootbecomesverylargeduetothederivativeeffect.Tz improvestherisetime,since1 + Tz s isa
derivativecontrolorahighpassfilter.
550
544UnitStepResponses
1
Theeffectofaddingthepoleat s = toG(s)istoincreasetherisetimeandtheovershoot.Thesystemis
Tp
lessstable.WhenT p > 0.707 ,theclosedloopsystemisstable.
551
5-45) You may use the ACSYS software developed for this book. For description refer to Chapter 9. We
use a MATLAB code similar to toolbox 2-2-1 and those in Chapter 5 to solve this problem.
(a) Using Toolbox 5-9-3

clear all
num = [];
den = [0 -0.55 -1.5];
G=zpk(num,den,1)
t=0:0.001:15;
step(G,t);
hold on;
for Tz=[1 5 20];
t=0:0.001:15;
num = [-1/Tz];
den = [0 -0.55 -1.5];
G=zpk(num,den,1)
step(G,t);
hold on;
end
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
Zero/pole/gain:
1
------------------
s (s+0.55) (s+1.5)
Zero/pole/gain:
(s+1)
------------------
s (s+0.55) (s+1.5)
Zero/pole/gain:
(s+0.2)
------------------
s (s+0.55) (s+1.5)
Zero/pole/gain:
(s+0.05)
------------------
s (s+0.55) (s+1.5)
552
(b)
clear all
for Tz=[0 1 5 20];
t=0:0.001:15;
num = [Tz 1];
den = [1 2 2];
G=tf(num,den)
step(G,t);
hold on;
end
ylabel('y(t)')
Transfer function:
1
-------------
s^2 + 2 s + 2
Transfer function:
s + 1
-------------
s^2 + 2 s + 2
Transfer function:
5 s + 1
-------------
s^2 + 2 s + 2
553
Transfer function:
20 s + 1
-------------
s^2 + 2 s + 2
Follow the same procedure for other parts.
5-46) Since the system is linear we use superposition to find Y, for inputs X and D
First, consider D = 0
kK
Js + a
Kf
Then
;
1
554
Kk
s ( Js + a + KK f k )

According to above block diagram:
1
1 1
Now consider X = 0, then:
k
K Kf
+1
Js + a s
Accordingly,
and
In this case, E(s) = D-KY(s)

1
1 1
555
Now the steady state error can be easily calculated by:

1
1/ : lim
1
1
lim lim 0
1
1 1
: lim
1
1
lim
1
lim
(c) The overall response is obtained through superposition
Y ( s) = Y ( s) D =0 + Y ( s ) X =0
y (t ) = y (t ) d (t ) =0 + y (t ) x (t ) =0
MATLAB
clear all
syms s K k J a Kf
X=1/s;
D=1/s^2
Y=K*k*X/(s*(J*s+a+K*k*Kf)+K*k)+k*(Kf+s)*D/(s*(J*s+a)+k*(Kf+s)*K)
ilaplace(Y)
D =
1/s^2
Y =
K*k/s/(s*(J*s+a+K*k*Kf)+K*k)+k*(Kf+s)/s^2/(s*(J*s+a)+k*(Kf+s)*K)
ans =
1+t/K+1/k/K^2/Kf/(a^2+2*a*K*k+K^2*k^2-
4*J*K*k*Kf)^(1/2)*sinh(1/2*t/J*(a^2+2*a*K*k+K^2*k^2-4*J*K*k*Kf)^(1/2))*exp(-
1/2*(a+K*k)/J*t)*(a^2+a*K*k-2*J*K*k*Kf)-cosh(1/2*t/J*(a^2+2*a*K*k*Kf+K^2*k^2*Kf^2-
4*J*K*k)^(1/2))*exp(-1/2*(a+K*k*Kf)/J*t)-(a+K*k*Kf)/(a^2+2*a*K*k*Kf+K^2*k^2*Kf^2-
4*J*K*k)^(1/2)*sinh(1/2*t/J*(a^2+2*a*K*k*Kf+K^2*k^2*Kf^2-4*J*K*k)^(1/2))*exp(-
1/2*(a+K*k*Kf)/J*t)+1/k/K^2/Kf*a*(-1+exp(-
1/2*(a+K*k)/J*t)*cosh(1/2*t/J*(a^2+2*a*K*k+K^2*k^2-4*J*K*k*Kf)^(1/2)))
556

5-47)(a)Findthe e(t )dt whene(t)istheerrorintheunitstepresponse.
0

Asthesystemisstablethen 0
e(t )dt willconvergetoaconstantvalue:
E ( s)

0
e(t )dt = lim s
s 0 s
= lim E ( s )
s 0
Y ( s) G ( s) ( A s + 1)( A2 s + 1)...( An s + 1)
= = 1 nm
X ( s) 1 + G ( s) ( B1 s + 1)( B2 s + 1)...( Bm s + 1)
Try to relate this to Equation (5-40).
G ( s) 1
E ( s ) = X ( s) Y ( s) = X (s) X (s) = X (s)
1 + G ( s) 1 + G (s)
( A s + 1)( A2 s + 1)...( An s + 1) ( B1s + 1)( B2 s + 1)...( Bm s + 1) ( A1s + 1)( A2 s + 1)...( An s + 1)
= 1 1 X (s) = X (s)
( B1s + 1)( B2 s + 1)...( Bm s + 1) ( B1s + 1)( B2 s + 1)...( Bm s + 1)
1 ( B s + 1)( B2 s + 1)...( Bm s + 1) ( A1s + 1)( A2 s + 1)...( An s + 1)
lim E ( s ) = lim 1
s 0 s 0 s
( B1s + 1)( B2 s + 1)...( Bm s + 1)
= ( B1 + B2 + L + Bm ) ( A1 + A2 + L + An )
( B1s + 1)( B2 s + 1)...( Bm s + 1)

G ( s) = 1
( B1s + 1)( B2 s + 1)...( Bm s + 1) ( A1s + 1)( A2 s + 1)...( An s + 1)
( A1s + 1)( A2 s + 1)...( An s + 1)
G ( s) =
( B1s + 1)( B2 s + 1)...( Bm s + 1) ( A1s + 1)( A2 s + 1)...( An s + 1)
1 ( A s + 1)( A2 s + 1)...( An s + 1)
ess = lim s 1 1 =0
s 0 s ( B1s + 1)( B2 s + 1)...( Bm s + 1)
557
1 1
(b)Calculate =
K lim sG ( s )
s 0
Recall
G ( s) 1
E ( s ) = X ( s) Y ( s) = X ( s) X (s) = X (s)
1 + G(s) 1 + G(s)
Hence

1 1 1 1
lim E ( s ) = lim X ( s ) = lim = lim =
s 0 s 0 1 + G ( s ) s 0 s + sG ( s ) s 0 sG ( s ) Kv
Ramp Error Constant
558
5-48)
10 10
25 10 35 25 10
Comparing with the second order prototype system and matching denominators:
25 10
35 2
exp 0.25 1.386 0.210

1 1
3.2
0.1 80, 0 0.69
Let = 0.4 and n = 80
Then
2 35 29
56.25
clear all
p=29;
K=56.25;
num = [10 10*K];
den = [1 35+p 25*p+10*K];
G=tf(num,den)
step(G);
ylabel('y(t)')
Transfer function:
10 s + 562.5
-----------------
s^2 + 64 s + 1288
559
5-49) According to the maximum overshoot:
which should be less than t, then
or 1
/t
n 1 2
/t n 1 2
560
5-50) Using a 2nd order prototype system format, from Figure 5-15, n is the radial distance from the
complex conjugate roots to the origin of the s-plane, then n with respect to the origin of the
shown region is n 3.6.
Therefore the natural frequency range in the region shown is around 2.6 n 4.6
On the other hand, the damping ratio at the two dashed radial lines is obtained from:
cos /2 sin
cos /2 sin
The approximation from the figure gives:
0.56
0.91
Therefore 0.56 0.91
b)
As Kp=2, then:
2
2 1 2
If the roots of the characteristic equations are assumed to be lied in the centre of the shown
region:
3 2
6 13 0
3 2
Comparing with the characteristic equation:
2 1 6 2
2 13 3.25
c) The characteristic equation
2 0
is a second order polynomial with two roots. These two roots can be determined by two terms
2(p+KKp) and KKpKI which includes four parameters. Regardless of the p and Kp values, we can
always choose K and so that to place the roots in a desired location.
561
5-51) a)
By substituting the values:
0.2
0.109
b) Speed of the motor is
0.2
0.109
.
lim 10 19.23
.
(c)
0.2
0.109
d)
0.2
0.109
0.2
0.109
As a result:
0.2
0.109 0.2
e) As exp 0.2 , then, = 0.404

.
According to the transfer function, 2 0.109, then 0.14 /
.
where 0.2 K<0.0845
. . .
f) As , then 0.6,
as 0.2 , therefore; 1
562
g) MATLAB
clear all
for K=[0.5 1 2];
t=0:0.001:15;
num = [0.2*K];
den = [1 0.109 0.2*K];
G=tf(num,den)
step(G,t);
hold on;
end
ylabel('y(t)')
Transfer function:
0.1
-------------------
s^2 + 0.109 s + 0.1
Transfer function:
0.2
-------------------
s^2 + 0.109 s + 0.2
Transfer function:
0.4
-------------------
s^2 + 0.109 s + 0.4
563
Rise time decreases with K increasing.
Overshoot increases with K.
564
5-52) exp 0.1, therefore, = 0.59
. .
As , then, 3.62
. .
1 3
3 3
Therefore:
3 3 2
If p is a non-dominant pole; therefore comparing both sides of above equation and:
2 3
2 3
If we consider p = 10a (non-dominant pole), = 0.6 and n = 4, then:
4.8 10 3
48 16 3
160 160
8.3
89.8
83
clear all
K=160;
a=8.3;
b=89.8;
p=83;
num = [K K*a];
den = [1 3+b 3*b+K K*a];
G=tf(num,den)
step(G);
ylabel('y(t)')
Transfer function:
160 s + 1328
-------------------------------
s^3 + 92.8 s^2 + 429.4 s + 1328
565
Both Overshoot and settling time values are met. No need to adjust parameters.
5-53) For the controller and the plant to be in series and using a unity feedback loop we have:
MATLAB
USE toolbox 5-8-3
clearall
num=[123];
denom=[3+sqrt(940)3sqrt(940)0.02+sqrt(.004.07)0.02sqrt(.004.07)10];
G=zpk(num,denom,60)
rlocus(G)
Zero/pole/gain:
60(s+1)(s+2)(s+3)
(s+10)(s^2+0.04s+0.0664)(s^2+6s+40)
566

Notethesystemhastwodominantcomplexpolesclosetotheimaginaryaxis.Letszoomintherootlocusdiagramand
usethecursortofindtheparametervalues.
567
AsshownforK=0.933thedominantclosedlooppolesareat0.46j0.626ANDOVERSHOOTIS
ALMOST10%.
IncreasingKwillpushthepolesclosertowardslessdominantzerosandpoles.Asaprocessthedesignprocessbecomesless
trivialandmoredifficult.
Toconfirmuse
M=feedback(G*.933,1)%Seetoolbox542
step(M)
Zero/pole/gain:
55.98(s+3)(s+2)(s+1)
(s+7.048)(s^2+0.9195s+0.603)(s^2+8.072s+85.29)
568
Toreducerisetime,thepoleshavetomovetolefttomakethesecondarypolesmoredominant.As
aresultthelittlebumpinthelefthandsideoftheabovegraphshouldrise.TryK=3:
>>M=feedback(G*3,1)
Zero/pole/gain:
180(s+3)(s+2)(s+1)
(s+5.01)(s^2+1.655s+1.058)(s^2+9.375s+208.9)
>>step(M)
**Try a higher K value, but looking at the root locus and the time plots, it appears that the
overshoot and rise time criteria will never be met simultaneously.
569
K=5
M=feedback(G*5,1) %See toolbox 5-4-2

step(M)
Zero/pole/gain:
300 (s+3) (s+2) (s+1)
---------------------------------------------------------
(s+4.434) (s^2 + 1.958s + 1.252) (s^2 + 9.648s + 329.1)
570
571
5-54) ForwardpathTransferFunction:
M ( s) K
G( s) = = 3 2

1 M ( s) s + ( 20 + a ) s + ( 200 + 20a ) s + 200a K
Fortype1system, 200a K = 0 ThusK=200a
Ramperrorconstant:
K 200 a
K v = lim sG ( s ) = = = 5 Thusa=10 K=2000
s 0 200 + 20a 200 + 20 a
MATLABSymbolictoolcanbeusedtosolveabove.Weuseittofindtherootsforthenextpart:
>>symssaK
>>solve(5*200+5*20*a200a)
ans=
10
>>D=(s^2+20*s+200)*s+a))
D=
(s^2+20*s+200)*(s+a)
>>expand(D)
ans=
s^3+s^2*a+20*s^2+20*s*a+200*s+200*a
>>solve(ans,s)
ans=
10+10*i
1010*i
Theforwardpathtransferfunctionis Thecontrollertransferfunctionis
572
G ( s ) =
2000
Gc ( s ) =
G(s)
=
(
20 s + 10 s + 100
2
)
(
s s + 30 s + 400
2
) Gp (s) (s 2
+ 30 s + 400 )
Themaximumovershootoftheunitstepresponseis0percent.
MATLAB
clear all
K=2000;
a=10;
num = [];
den = [-10+10i -10-10i -a];
G=zpk(num,den,K)
step(G);
ylabel('y(t)')
Zero/pole/gain:
2000
-------------------------
(s+10) (s^2 + 20s + 200)
Clearly PO=0.
573
5-55)
ForwardpathTransferFunction:
M ( s) K
G( s) = = 3 2

1 M ( s) s + ( 20 + a ) s + ( 200 + 20a ) s + 200a K
Ramperrorconstant:
K 200 a
K v = lim sG ( s ) = = = 9 Thusa=90 K=18000
s 0 200 + 20a 200 + 20 a
MATLABSymbolictoolcanbeusedtosolveabove.Weuseittofindtherootsforthenextpart:
>>symssaK
solve(9*200+9*20*a200*a)
ans=
90
>>D=(s^2+20*s+200)*s+a))
D=
(s^2+20*s+200)*(s+a)
>>expand(D)
ans=
s^3+s^2*a+20*s^2+20*s*a+200*s+200*a
>>solve(ans,s)
ans=
10+10*i
1010*i
574
G ( s ) =
18000
Gc ( s ) =
G(s)
=
(
180 s + 10 s + 100
2
)
(
s s + 110 s + 2000
2
) Gp ( s) ( s + 110s + 2000 )
2
Themaximumovershootoftheunitstepresponseis4.3percent.
Fromtheexpressionfortheramperrorconstant,weseethatasaorKgoestoinfinity, Kv approaches10.
Thusthemaximumvalueof Kv thatcanberealizedis10.ThedifficultieswithverylargevaluesofKand
aarethatahighgainamplifierisneededandunrealisticcircuitparametersareneededforthecontroller.
clear all
K=18000;
a=90;
num = [];
den = [-10+10i -10-10i -a];
G=zpk(num,den,K)
step(G);
ylabel('y(t)')
Zero/pole/gain:
18000
-------------------------
(s+90) (s^2 + 20s + 200)
PO is less than 4.
575
5-56)(a)RamperrorConstant:
MATLAB
clearall
symssKpKdkv
Gnum=(Kp+Kd*s)*1000
Gden=(s*(s+10))
G=Gnum/Gden
Kv=s*G
s=0
eval(Kv)
Gnum=
1000*Kp+1000*Kd*s
Gden=
s*(s+10)
G=
(1000*Kp+1000*Kd*s)/s/(s+10)
Kv=
(1000*Kp+1000*Kd*s)/(s+10)
s=
0
ans=
100*Kp
1000 ( K P + K D s ) 1000 K P
K v = lim s = = 100 K P = 1000 Thus K P = 10
s 0 s ( s + 10) 10
Kp=10
clears
symss
Mnum=(Kp+Kd*s)*1000/s/(s+10)
Mden=1+(Kp+Kd*s)*1000/s/(s+10)
576
Kp=
10
Mnum=
(10000+1000*Kd*s)/s/(s+10)
Mden=
1+(10000+1000*Kd*s)/s/(s+10)
ans=
(s^2+10*s+10000+1000*Kd*s)/s/(s+10)
s + (10 + 1000 K D ) s + 1000 K P = 0

2
CharacteristicEquation:
Matchwitha2ndorderprototypesystem
n = 1000 K P = 10000 = 100 rad/sec 2 n = 10 + 1000 K D = 2 0.5 100 = 100
solve(10+1000*Kd100)
ans=
9/100
90
Thus KD = = 0.09
1000
Usethesameprocedureforotherparts.
(b)For K v = 1000 and = 0.707 ,andfrompart(a), n = 100 rad/sec,
131.4
2 n = 10 + 1000 K D = 2 0.707 100 = 141.4 Thus KD = = 0.1314
1000
(c)For K v = 1000 and = 1.0 ,andfrompart(a), n = 100 rad/sec,
190
2 n = 10 + 1000 K D = 2 1 100 = 200 Thus KD = = 0.19
1000
577
5-57)Theramperrorconstant:
1000 ( K P + K D s )
K v = lim s = 100 K P = 10, 000 Thus K P = 100
s 0 s ( s + 10)
1000 (100 + K D s )
Theforwardpathtransferfunctionis: G(s) =
s ( s + 10)
clear all
for KD=0.2:0.2:1.0;
num = [-100/KD];
den = [0 -10];
G=zpk(num,den,1000);
M=feedback(G,1)
step(M);
hold on;
end
ylabel('y(t)')
Zero/pole/gain:
1000(s+500)
(s^2+1010s+5e005)
Zero/pole/gain:
1000(s+250)
(s+434.1)(s+575.9)
Zero/pole/gain:
1000(s+166.7)
(s+207.7)(s+802.3)
578
Zero/pole/gain:
1000(s+125)
(s+144.4)(s+865.6)
Zero/pole/gain:
1000(s+100)
(s+111.3)(s+898.7)
UsethecursortoobtainthePOandtrvalues.
ForpartbthemaximumvalueofKDresultsintheminimumovershoot.
579
5-58)(a)ForwardpathTransferFunction:
4500 K ( K P + K D s )
G ( s ) = Gc ( s )G p ( s ) =
s ( s + 361.2)
4500 KK P
RampErrorConstant: K v = lim sG ( s ) = = 12.458 KK P
s 0 361.2
1 0.0802
e ss = = 0.001Thus KK P 80.2 Let K P = 1 and K = 80.2
Kv KK P
clear all
KP=1;
K=80.2;
figure(1)
num = [-KP];
den = [0 -361.2];
G=zpk(num,den,4500*K)
M=feedback(G,1)
step(M)
hold on;
for KD=00.0005:0.0005:0.002;
num = [-KP/KD];
den = [0 -361.2];
G=zpk(num,den,4500*K*KD)
M=feedback(G,1)
step(M)
end
ylabel('y(t)')
Zero/pole/gain:
360900 (s+1)
s (s+361.2)
Zero/pole/gain:
360900 (s+1)
(s+0.999) (s+3.613e005)
Zero/pole/gain:
180.45 (s+2000)
s (s+361.2)
Zero/pole/gain:
180.45 (s+2000)
580
(s^2 + 541.6s + 3.609e005)
Zero/pole/gain:
360.9 (s+1000)
s (s+361.2)
Zero/pole/gain:
360.9 (s+1000)
(s^2 + 722.1s + 3.609e005)
Zero/pole/gain:
541.35 (s+666.7)
s (s+361.2)
Zero/pole/gain:
541.35 (s+666.7)
(s^2 + 902.5s + 3.609e005)
Zero/pole/gain:
721.8 (s+500)
s (s+361.2)
Zero/pole/gain:
721.8 (s+500)
(s^2 + 1083s + 3.609e005)
581
KD tr(sec) ts (sec) MaxOvershoot(%)
0 0.00221 0.0166 37.1
0.0005 0.00242 0.00812 21.5
0.0010 0.00245 0.00775 12.2
0.0015 0.0024 0.0065 6.4
0.0016 0.00239 0.00597 5.6
0.0017 0.00238 0.00287 4.8
0.0018 0.00236 0.0029 4.0
0.0020 0.00233 0.00283 2.8
582
5-59)TheforwardpathTransferFunction:N=20
200 ( K P + K D s )
G(s) =
s ( s + 1)( s + 10)
Tostabilizethesystem,wecanreducetheforwardpathgain.Sincethesystemistype1,reducingthe
gaindoesnotaffectthesteadystateliquidleveltoastepinput.Let K P = 0.05
200 ( 0.05 + K D s )
G(s) =
s ( s + 1)( s + 10)
ALSO try other Kp values and compare your results.
clear all
figure(1)
KD=0
num = [];
den = [0 -1 -10];
G=zpk(num,den,200*0.05)
M=feedback(G,1)
step(M)
hold on;
for KD=0.01:0.01:0.1;
KD
num = [-0.05/KD];
G=zpk(num,den,200*KD)
M=feedback(G,1)
step(M)
end
ylabel('y(t)')
KD=
0
Zero/pole/gain:
10
s(s+1)(s+10)
Zero/pole/gain:
10
(s+10.11)(s^2+0.8914s+0.9893)
KD=
0.0100
583
Zero/pole/gain:
2(s+5)
s(s+1)(s+10)
Zero/pole/gain:
2(s+5)
(s+9.889)(s^2+1.111s+1.011)
KD=
0.0200
Zero/pole/gain:
4(s+2.5)
s(s+1)(s+10)
Zero/pole/gain:
4(s+2.5)
(s+9.658)(s^2+1.342s+1.035)
KD=
0.0300
Zero/pole/gain:
6(s+1.667)
s(s+1)(s+10)
Zero/pole/gain:
6(s+1.667)
(s+9.413)(s^2+1.587s+1.062)
KD=
0.0400
Zero/pole/gain:
8(s+1.25)
s(s+1)(s+10)
Zero/pole/gain:
8(s+1.25)
(s+9.153)(s^2+1.847s+1.093)
584
KD=
0.0500
Zero/pole/gain:
10(s+1)
s(s+1)(s+10)
Zero/pole/gain:
10(s+1)
(s+8.873)(s+1.127)(s+1)
KD=
0.0600
Zero/pole/gain:
12(s+0.8333)
s(s+1)(s+10)
Zero/pole/gain:
12(s+0.8333)
(s+8.569)(s+1.773)(s+0.6582)
KD=
0.0700
Zero/pole/gain:
14(s+0.7143)
s(s+1)(s+10)
Zero/pole/gain:
14(s+0.7143)
(s+8.232)(s+2.221)(s+0.547)
KD=
0.0800
Zero/pole/gain:
16(s+0.625)
s(s+1)(s+10)
Zero/pole/gain:
16(s+0.625)
(s+7.85)(s+2.673)(s+0.4765)
585
KD=
0.0900
Zero/pole/gain:
18(s+0.5556)
s(s+1)(s+10)
Zero/pole/gain:
18(s+0.5556)
(s+7.398)(s+3.177)(s+0.4255)
KD=
0.1000
Zero/pole/gain:
20(s+0.5)
s(s+1)(s+10)
Zero/pole/gain:
20(s+0.5)
(s+0.3861)(s+3.803)(s+6.811)
586
UnitstepResponseAttributes:
KD ts(sec) MaxOvershoot(%)
0.01 5.159 12.7
0.02 4.57 7.1
0.03 2.35 3.2
0.04 2.526 0.8
0.05 2.721 0
0.06 3.039 0
0.10 4.317 0
When K D = 0.05 therisetimeis2.721sec,andthestepresponsehasnoovershoot.
587
5-60) (a)For e ss = 1,
200 ( K P + K D s )
K v = lim sG ( s ) = lim s = 20 K P = 1 Thus K P = 0.05
s 0 s0 s ( s + 1)( s + 10)
200 ( 0.05 + K D s )
G ( s ) =
s ( s + 1)( s + 10)
Because of the choice of Kp this is the same as previous part.
5-61)
(a)ForwardpathTransferFunction:

100 K P +
KI
100 ( K P s + K I )
G ( s ) =
s For K v = 10, K v = lim sG ( s ) = lim s = K I = 10
s + 10 s + 100
2
s 0 s 0
(
s s + 10 s + 100
2
)

Thus K I = 10.
100 ( K P s + K I ) 100 ( K P s + 10 )
Gcl ( s ) = =
s + 10 s + 100 s + 100 ( K P s + K I )
2 2
s + 10 s + 100(1 + K P ) s + 1000
3 3
(b)Letthecomplexrootsofthecharacteristicequationbewrittenas s = + j15 and s = j15.
Thequadraticportionofthecharacteristicequationis
2
(
s + 2 s + + 225 = 0
2
)
s + 10 s + (100 + 100 K P ) s + 1000 = 0
3 2
Thecharacteristicequationofthesystemis
Thequadraticequationmustsatisfythecharacteristicequation.Usinglongdivisionandsolveforzero
remaindercondition.
588
s + (10 2 )
s + 2 s + + 225 s + 10 s + (100 + 100 K P ) s + 1000

2 2 3 2
s + 2 s + + 225 s
3 2
( 2
)
(
(10 2 ) s + 100 K P 125 s + 1000
2
) 2

(10 2 ) s + ( 20 4 ) s + (10 2 ) ( s + 225 )
2 2 2
(100 K P )
+ 3 20 125 s + 2 10 + 450 1250
2 3 2
3 2
Forzeroremainder, 2 10 + 450 1250 = 0 (1)
2
and 100 K P + 3 20 125 = 0 (2)
TherealsolutionofEq.(1)is = 2.8555 .FromEq.(2),
2
125 + 20 3
KP = = 15765
.
100
Thecharacteristicequationrootsare: s = 2.8555 + j15, 2.8555 j15, and s = 10 + 2 = 4.289
(c)RootContours:
Dividingbothsidesof s + 10 s + (100 + 100 K P ) s + 1000 = 0 bythetermsthatdonotcontainKpwehave:

3 2
100 K P s
1+ = 1 + Geq
s + 10 s + 100 s + 1000
3 2
100 K P s 100 K P s
Geq ( s ) = =
s + 10 s + 100 s + 1000
3 2
( s + 10 ) ( s 2
+ 100 )
RootContours:SeeChapter9toolbox952formoreinformation
clear all
Kp =.001;
num = [100*Kp 0];
den = [1 10 100 1000];
rlocus(num,den)
589
590
5-62) (a)ForwardpathTransferFunction:

100 K P +
KI
100 ( K P s + K I )
G ( s ) =
s For K = 10, K v = lim sG ( s ) = lim s = K I = 10
s + 10 s + 100
2 v
s 0 s 0
(
s s + 10 s + 100
2
)
Thustheforwardpathtransferfunctionbecomes
100 (10 + K P s )
G(s) =
(
s s + 10 s + 100
2
)
100 ( K P s + K I ) 100 ( K P s + 10 )
Gcl ( s ) = =
s + 10 s + 100 s + 100 ( K P s + K I )
2 2
s + 10 s + 100(1 + K P ) s + 1000
3 3
clear all
for Kp=.4:0.4:2;
num = [100*Kp 1000];
den =[1 10 100 0];
[numCL,denCL]=cloop(num,den);
GCL=tf(numCL,denCL);
step(GCL)
hold on;
end
591
Usethecursertofindthemaximumovershootandrisetime.ForexamplewhenKp=2,PO=43andtr100%=0.152
sec.
Transferfunction:
200s+1000
s^3+10s^2+300s+1000
592
5-63)
100 ( K P s + K I )
G(s) =
(
s s + 10 s + 100
2
)
For K v = 100,
100 ( K P s + K I )
K v = lim sG ( s ) = lim s = K I = 100 Thus K I = 100.
s 0 s 0
(
s s + 10 s + 100
2
)
s + 10 s + (100 + 100 K P ) s + 100 K I = 0
3 2
(b)Thecharacteristicequationis
RouthTabulation:
3
s 1 100 + 100 K P
2
s 10 10,000
Forstability, 100 K P 900 > 0 Thus KP > 9
1
s 100 K P 900 0
0
s 10,000
7. ActivateMATLAB
9. Typein
Acsys
10. ThenpressthetransferfunctionSymbolicandentertheCharacteristicequation
11. ThenpresstheRouthHurwitzbutton
593

RH=
[1,100+100*kp]
[10,10000]
[900+100*kp,0]
[(9000000+1000000*kp)/(900+100*kp),0]
594
RootContours:
100 K P s 100 K P s
Geq ( s ) = 3 2
=
s + 10 s + 100 s + 10,000 ( s + 23.65)( s 6.825 + j19.4)( s 6.825 j19.4)
RootContours:SeeChapter9toolbox952formoreinformation
clear all
Kp =.001;
num = [100*Kp 0];
den = [1 10 100 10000];
rlocus(num,den)
595
(c) K I = 100
100 ( K P s + 100 )
G(s) =
(
s s + 10 s + 100
2
)
Thefollowingmaximumovershootsofthesystemarecomputedforvariousvaluesof K P .
clear all
Kp=[15 20 22 24 25 26 30 40 100 1000];
[N,M]=size(Kp);
for i=1:M
num = [100*Kp(i) 10000];
den = [1 10 100 0];
GCL=tf(numCL,denCL);
figure(i)
step(GCL)
end
KP 15 20 22 24 25 26 30 40 100 1000
596
ymax 1.794 1.779 1.7788 1.7785 1.7756 1.779 1.782 1.795 1.844 1.859
WhenKP=25,minimumymax=1.7756
Use: close all to close all the figure windows.
597
5-64) MATLAB solution is the same as 5-63.
100 ( K P s + K I ) 100 K I
G(s) = For K v = = 10, K I = 10
(
s s + 10 s + 100
2
) 100
(b)CharacteristicEquation: ( )
s3 + 10 s 2 + 100 K + 1 s + 1000 = 0
P
RouthTabulation:
3
s 1 100 + 100 K P
2
s 10 1000
Forstability,KP>0
1
s 100 K P 0
0
s 1000
RootContours:
100 K P s
Geq ( s ) = 3 2

s + 10 s + 100 s + 1000
598
(c)Themaximumovershootsofthesystemfordifferentvaluesof K P rangingfrom0.5to20are
computedandtabulatedbelow.
KP 0.5 1.0 1.6 1.7 1.8 1.9 2.0 3.0 5.0 10 20
ymax 1.393 1.275 1.2317 1.2416 1.2424 1.2441 1.246 1.28 1.372 1.514 1.642
WhenKP=1.7,maximumymax=1.2416
5-65)
KDs + KPs + KI KI 2
2
KI
Gc ( s ) = K P + K D s + = = (1 + K D 1 s ) K P 2 +
s s s
where
K P = K P 2 + K D1 K I 2 K D = K D1 K P 2 KI = KI 2
G ( s ) = Gc ( s )G p ( s ) =
100 (K D s2 + KP s + KI )
(
s s + 10 s + 100
2
)
Andrenametheratios: K D / K P = A, KI / KP = B
Thus
KI
K v = lim sG ( s ) = 100 = 100
s 0 100
K I = 100
599
ForKDbeingsufficientlysmall:
ForwardpathTransferFunction: CharacteristicEquation:
100 ( K P s + 100 )
G ( s ) = s + 10 s + (100 + 100 K P ) s + 10, 000 = 0
3 2

(
s s + 10 s + 100
2
)
Forstability,Kp>9.SelectKp=10andobservetheresponse.
clear all
Kp=10;
num = [100*Kp 10000];
den =[1 10 100 0];
GCL=tf(numCL,denCL)
step(GCL)
Transfer function:
1000 s + 10000
-----------------------------
s^3 + 10 s^2 + 1100 s + 10000
ObviouslybyincreasingKpmoreoscillationswilloccur.AddKDtoreduceoscillations.
clear all
Kp=10;
Kd=2;
num = [100*Kd 100*Kp 10000];
den =[1 10 100 0];
GCL=tf(numCL,denCL)
step(GCL)
5100
Transfer function:
200 s^2 + 1000 s + 10000
------------------------------
s^3 + 210 s^2 + 1100 s + 10000
UnitstepResponse
Therisetimeseemsreasonable.ButweneedtoincreaseKptoimproveapproachtosteadystate.
IncreaseKptoKp=30.
clear all
Kp=30;
Kd=1;
num = [100*Kd 100*Kp 10000];
den =[1 10 100 0];
GCL=tf(numCL,denCL)
step(GCL)
Transfer function:
100 s^2 + 3000 s + 10000
------------------------------
s^3 + 110 s^2 + 3100 s + 10000
5101
To obtain a better response continue adjusting KD and KP.
5102
5-66) This problem has received extended treatment in Chapter 6, Control Lab see Section 6-6.
For the sake simplicity, this problem we assume the control force f(t) is applied in parallel to the
spring K and damper B. We will not concern the details of what actuator or sensors are used.
Lets look at Figure 4-84 and equations 4-322 and 4-323.
xc
mc
mc
ks cs
ks cs x
xw m
mw
mw k c
kw kw cw
y y
cw
(a) (b) (c)
Figure 4-84 Quarter car model realization: (a) quarter car, (b) 2 degree of
freedom, and (c) 1 degree of freedom model.
Theequationofmotionofthesystemisdefinedasfollows:
mx&&(t ) + cx& (t ) + kx (t ) = cy& (t ) + ky (t ) (4-322)
whichcanbesimplifiedbysubstitutingtherelationz(t)=x(t)y(t)andnondimensionalizingthecoefficientsto
theform
z (t ) + 2 n z& (t ) + n 2 z (t ) = &&
&& y (t ) (4-323)
TheLaplacetransformofEq.(4323)yieldstheinputoutputrelationship
5103
Z (s) 1
= 2 (4-324)
Y&&( s ) s + 2 n s + n 2
Now lets apply control see section 6-6 for more detail.
x
m
k c
k f(t)
y
For simplicity and better presentation, we have scaled the control force as kf(t) we rewrite (4-324) as:
mx&&(t ) + cx& (t ) + kx(t ) = cy& (t ) + ky (t ) + kf (t )

z (t ) + 2n z& (t ) + n 2 z (t ) = &&
&& y (t ) + n 2 f (t )
(4-324)
s 2 + 2n s + n 2 = A( s ) + n 2 F ( s )
A( s ) = Y&&( s )
Setting the controller structure such that the vehicle bounce Z ( s ) = X ( s ) Y ( s ) is minimized:
K
F (s) = 0 K P + K D s + I Z (s)
s
Z (s) 1
=
A( s ) K
s 2 + 2n s + n 2 1 + K P + K D s + I
s
Z (s) s
= 3
A( s ) s + 2n s + n ( (1 + K P ) s + K D s 2 + K I )
2 2
See Equation (6-4).
For proportional control KD=KI=0.
5104
Pick = 0.707 and n = 1 for simplicity. This is now an underdamped system.
Use MATLAB to obtain response now.

clear all
Kp=1;
Kd=0;
Ki=0;
num = [-1 0];
den =[1 2*0.707+Kd 1+Kp Ki];
G=tf(num,den)
step(G)
Transfer function:
-s
---------------------
s^3 + 1.414 s^2 + 2 s
Adjust parameters to get the desired response if necessary.
The process is the same for parts b, c and d.
5105
5-67) Replace F(s) with
K
F ( s ) = X ref K P + K D s + I X ( s )
s
B
2n =
M
K
n 2 =
M
X ( s) 1
=
X ref ( s ) K
s 2 + 2n s + n 2 + K P + K D s + I
s
Use MATLAB to obtain response now.

clear all
Kp=1;
Kd=0;
Ki=0;
B=10;
K=20;
M=1;
omega=sqrt(K/M);
zeta=(B/M)/2/omega;
num = [1 0];
den =[1 2*zeta*omega+Kd omega^2+Kp Ki];
G=tf(num,den)
step(G)
ransfer function:
s
-------------------
s^3 + 10 s^2 + 21 s
T0 achieve the proper response, adjust controller gains accordingly.
5106
5-68) From problem 4-3
a) Rotational kinetic energy:
Translational kinetic energy:
Relation between translational displacement and rotational displacement:
1
2
Potential energy:
As we know , then:
1 1 1
2 2 2
By differentiating, we have:
0
Since cannot be zero, then 0
b)
c)
5107
1 1 1
2 2 2
where at the maximum energy.
1 1
2 2
Then:
1 1
2 2
Or:
J
d) G (s) =
(ms + K )
2

K=100;
J=5;
m=25;
G=tf([J],[m 0 K])
Pole(G)
impulse(G,10)
Transfer function:
5
------------
25 s^2 + 100
ans =
0 + 2.0000i
0 - 2.0000i
5108
Uncontrolled
With a proportional controller one can adjust the oscillation amplitude the transfer function is
rewritten as:
JK p
Gcl ( s ) =
(ms + K + JK p )
2

Kp=0.1
K=100;
J=5;
m=25;
G=tf([J*Kp],[m 0 (K+J*Kp)])
Pole(G)
impulse(G,10)
5109
Kp =
0.1000
Transfer function:
0.5
--------------
25 s^2 + 100.5
ans =
0 + 2.0050i
0 - 2.0050i
A PD controller must be used to damp the

oscillation and reduce overshoot. Use Example
5-11-1 as a guide.
5110
5-69) From Problem 4-6 we have:
a)
y2 y1
K ( y1 y 2 )
K ( y1 y 2 )
mgy& 2 Mgy&1
b) From Newtons Law:
If y1 and y2 are considered as a position and v1 and v2 as velocity variables
Then:
The output equation can be the velocity of the engine, which means
c)
Obtaining requires solving above equation with respect to Y2(s)
From the first equation:
Substituting into the second equation:
5111
By solving above equation:
1
2
Replace Force F with a proportional controller so that F=K(Z-Zref):
KP
Zref_FZ
1
2
1
2 1
5-70) Also see derivations in 4-9.
x(t)
F
L
M
Here is an alternative representation including friction (damping) . In this case the angle is
measured differently.
Lets find the dynamic model of the system:
5112
1)
2)
Let . If is small enough then 1 and , therefore
which gives:
s
F s
Ignoring friction 0.
s
F s
where
Ignoring actuator dynamics (DC motor equations), we can incorporate feedback control using a
series PD compensator and unity feedback. Hence,
KP+KDs
R_F
F ( s) = K p ( R ( s ) ) K D s ( R ( s ) )
The system transfer function is:
A( K p + KDs)
= 2
R (s + K D s + A ( K p B )
Control is achieved by ensuring stability (Kp>B)
5113
Use Routh Hurwitz to establish stability first. Use Acsys to do that as

demonstrated in this chapter problems. Also Chapter 2 has many examples.
Use MATLAB to simulate response:

clear all
Kp=10;
Kd=5;
A=10;
B=8;
num = [A*Kd A*Kp];
den =[1 Kd A*(Kp-B)];
G=tf(num,den)
step(G)
Transfer function:
50 s + 100
--------------
s^2 + 5 s + 20
Adjust parameters to achieve desired response. Use THE PROCEDURE in Example 5-11-1.
You may look at the root locus of the forward path transfer function to get a better perspective.
5114
A ( K p + K D s ) AK D ( z + s )
= =
E s 2 AB s 2 AB
fix z and vary K D .
clear all
z=100;
Kd=0.01;
A=10;
B=8;
num = [A*Kd A*Kd*z];
den =[1 0 -(A*B)];
G=tf(num,den)
rlocus(G)
Transfer function:
0.1 s + 10
----------
s^2 80
Forz=10,alargeKD=0.805resultsin:
5115
clear all
Kd=0.805;
Kp=10*Kd;
A=10;
B=8;
num = [A*Kd A*Kp];
G=tf(num,den)
pole(G)
zero(G)
step(G)
Transfer function:
8.05 s + 80.5
-------------------
s^2 + 0.805 s + 0.5
ans =
-0.4025 + 0.5814i
-0.4025 - 0.5814i
ans =
-10
Looking at dominant poles we expect to see an oscillatory response with overshoot

close to desired values.
For a better design, and to meet rise time criterion, use Example 5-
11-1.
5116
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
Chapter 6 THE CONTROL LAB
Part 1) Solution to Lab questions within Chapter 6
6-4-1 Open Loop Speed
1. Open loop speed response using SIMLab:

a. +5 V input:
The form of response is like the one that we expected; a second order system response with overshoot and
oscillation. Considering an amplifier gain of 2 and K b = 0.1 , the desired set point should be set to 2.5 and as
seen in the figure, the final value is approximately 50 rad/sec which is armature voltage divided by K b . To find
the above response the systems parameters are extracted from:
Ra J m Ra J m k b k m m
m = , B= = 0.000792kg m 2 / sec
Ra B + k b k m Ra m
6-1
b. +15 V input:
c. 10 V input:
6-1
2. Study of the effect of viscous friction:
The above figure is plotted for three different friction coefficients (0, 0.001, 0.005) for 5 V armature input. As
seen in figure, two important effects are observed as the viscous coefficient is increased. First, the final steady
state velocity is decreased and second the response has less oscillation. Both of these effects could be predicted
from Eq. (5-114) by increasing damping ratio .
6-1
3. Additional load inertia effect:
As the overall inertia of the system is increased by 0.005 / 5.2 2 and becomes 1.8493 10 3 kg.m2, the
mechanical time constant is substantially increased and we can assume the first order model for the motor
(ignoring the electrical sub-system) and as a result of this the response is more like an exponential form. The
above results are plotted for 5 V armature input.
6-1
4. Reduce the speed by 50% by increasing viscous friction:
As seen in above figure, if we set B=0.0075 N.s/m the output speed drop by half comparing with the case that
B=0 N.s/m. The above results are plotted for 5 V armature input.
6-1
5. Study of the effect of disturbance:
Repeating experiment 3 for B=0.001 N.s/m and TL =0.05 N.m result in above figure. As seen, the effect of
disturbance on the speed of open loop system is like the effect of higher viscous friction and caused to decrease
the steady state value of speed.
6-1
6. Using speed response to estimate motor and load inertia:
Using first order model we are able to identify system parameters based on unit step response of the system. In
above plot we repeated the experiments 3 with B=0.001 and set point voltage equal to 1 V. The final value of the
speed can be read from the curve and it is 8.8, using the definition of system time constant and the cursor we read
63.2% of speed final value 5.57 occurs at 0.22 sec, which is the system time constant. Considering Eq. (5-116),
and using the given value for the rest of parameters, the inertia of the motor and load can be calculated as:
m ( Ra B + K m K b ) 0.22(1.35 0.001 + 0.01)

J= = = 1.8496 10 3 kg.m2
Ra 1.35
We also can use the open loop speed response to estimate B by letting the speed to coast down when it gets to the
steady state situation and then measuring the required time to get to zero speed. Based on this time and energy
conservation principle and knowing the rest of parameters we are able to calculate B. However, this method of
identification gives us limited information about the system parameters and we need to measure some parameters
directly from motor such as Ra , K m , K b and so on.
So far, no current or voltage saturation limit is considered for all simulations using SIMLab software.
6-1
7. Open loop speed response using Virtual Lab:

a. +5 V:
6-1
b. +15 V:
c. 10 V:
6-1
Comparing these results with the part 1, the final values are approximately the same but the shape of responses is
closed to the first order system behavior. Then the system time constant is obviously different and it can be
identified from open loop response. The effect of nonlinearities such as saturation can be seen in +15 V input with
appearing a straight line at the beginning of the response and also the effects of noise and friction on the response
can be observed in above curves by reducing input voltage for example, the following response is plotted for a 0.1
V step input:
6-1
6-1
8. Identifying the system based on open loop response:
Open loop response of the motor to a unit step input voltage is plotted in above figure. Using the definition of
time constant and final value of the system, a first order model can be found as:
9
G(s) = ,
0.23s + 1
where the time constant (0.23) is found at 5.68 rad/sec (63.2% of the final value).
6-1
6-4-2 Open Loop Sine Input
9. Sine input to SIMLab and Virtual Lab (1 V. amplitude, and 0.5, 5, and 50 rad/sec frequencies)
a. 0.5 rad/sec (SIMLab):
6-1
b. 5 rad/sec (SIMLab):
c. 50 rad/sec (SIMLab):
6-1
d. 0.5 rad/sec (Virtual Lab):
e. 5 rad/sec (Virtual Lab):
6-1
f. 50 rad/sec (Virtual Lab):
6-1
10. Sine input to SIMLab and Virtual Lab (20 V. amplitude)

a. 0.5 rad/sec (SIMLab):
b. 5 rad/sec (SIMLab):
6-1
c. 50 rad/sec (SIMLab):
d. 0.5 rad/sec (Virtual Lab):
6-1
e. 5 rad/sec (Virtual Lab):
6-1
f. 5 rad/sec (Virtual Lab):
In both experiments 9 and 10, no saturation considered for voltage and current in SIMLab software. If
we use the calculation of phase and magnitude in both SIMLab and Virtual Lab we will find that as
input frequency increases the magnitude of the output decreases and phase lag increases. Because of
existing saturations this phenomenon is more sever in the Virtual Lab experiment (10.f). In this
experiments we observe that M = 0.288 and = 93.82 o for = 50.
6-1
6-4-3 Speed Control

11. Apply step inputs (SIMLab)
In this section no saturation is considered either for current or for voltage.
a. +5 V:
6-1
b. +15 V:
c. -10 V:
6-1
6-1
12. Additional load inertia effect:

a. +5 V:
b. +15 V:
6-1
c. -10 V:
13. Study of the effect of viscous friction:
6-1
As seen in above figure, two different values for B are selected, zero and 0.0075. We could change the
final speed by 50% in open loop system. The same values selected for closed loop speed control but as
seen in the figure the final value of speeds stayed the same for both cases. It means that closed loop
system is robust against changing in systems parameters. For this case, the gain of proportional
controller and speed set point are 10 and 5 rad/sec, respectively.
14. Study of the effect of disturbance:
6-1
Repeating part 5 in section 6-4-1 for B=0.001 and TL =0.05 N.m result in above figure. As seen, the effect of
disturbance on the speed of closed loop system is not substantial like the one on the open loop system in part 5,
and again it is shown the robustness of closed loop system against disturbance. Also, to study the effects of
conversion factor see below figure, which is plotted for two different C.F. and the set point is 5 V.
6-1
By decreasing the C.F. from 1 to 0.2, the final value of the speed increases by a factor of 5.
15. Apply step inputs (Virtual Lab)

a. +5 V:
6-1
b. +15 V:
c. 10 V:
6-1
As seen the responses of Virtual Lab software, they are clearly different from the same results of
SIMLab software. The nonlinearities such as friction and saturation cause these differences. For
example, the chattering phenomenon and flatness of the response at the beginning can be considered as
some results of nonlinear elements in Virtual Lab software.
6-4-4 Position Control
16. 160 o step input (SIMLab)
6-1
6-1
17. 0.1 N.m step disturbance
18. Examine the effect of integral control
6-1
In above figure, an integral gain of 1 is considered for all curves. Comparing this plot with the previous
one without integral gain, results in less steady state error for the case of controller with integral part.
19. Additional load inertia effect (J=0.0019, B=0.004):
6-1
6-1
20. Set B=0:
21. Study the effect of saturation
6-1
The above figure is obtained in the same conditions of part 20 but in this case we considered 10 V. and
4 A. as the saturation values for voltage and current, respectively. As seen in the figure, for higher
proportional gains the effect of saturations appears by reducing the frequency and damping property of
the system.
22. Comments on Eq. 5-126
After neglecting of electrical time constant, the second order closed loop transfer function of position
control obtained in Eq. 5-126. In experiments 19 through 21 we observe an under damp response of a
second order system. According to the equation, as the proportional gain increases, the damped
frequency must be increased and this fact is verified in experiments 19 through 21. Experiments16
through 18 exhibits an over damped second order system responses.
23. In following, we repeat parts 16 and 18 using Virtual Lab:
Study the effect of integral gain of 5:
6-1
6-1
Ch. 6 Problem Solutions
Part 2) Solution to Problems in Chapter 6
6-1. In order to find the current of the motor, the motor constant has to be separated from the electrical
component of the motor.
The response of the motor when 5V of step input is applied is:
a) The steady state speed: 41.67rad/sec

b) It takes 0.0678 second to reach 63% of the steady state speed (26.25rad/sec). This is the time
constant of the motor.
c) The maximum current: 2.228A
6-1
11.2
The steady state speed at 5V step input is 50rad/sec.
a) It takes 0.0797 seconds to reach 63% of the steady state speed (31.5rad/sec).
b) The maximum current: 2.226A
c) 100rad/sec
6-1
6-3
a) 50rad/sec
b) 0.0795 seconds
c) 2.5A. The current
d) When Jm is increased by a factor of 2, it takes 0.159 seconds to reach 63% of its steady state
speed, which is exactly twice the original time period . This means that the time constant has
been doubled.
6-1
6-4
Part 1: Repeat problem 6-1 with TL = -0.1Nm
a) It changes from 41.67 rad/sec to 25 rad/sec.

b) First, the speed of 63% of the steady state has to be calculated.
41.67 - (41.67 - 25) 0.63 = 31.17 rad/sec.
The motor achieves this speed 0.0629 seconds after the load torque is applied
c) 2.228A. It does not change
a) It changes from 50 rad/sec to 30 rad/sec.

b) The speed of 63% of the steady state becomes
50 - (50 - 30) 0.63 = 37.4 rad/sec.
The motor achieves this speed 0.0756 seconds after the load torque is applied
c) 2.226A. It does not change.
6-1
a) It changes from 50 rad/sec to 30 rad/sec.

b) 50 - (50 - 30) 0.63 = 37.4 rad/s
The motor achieves this speed 0.0795 seconds after the load torque is applied. This is the same
as problem 6-3.
c) 2.5A. It does not change
d) As TL increases in magnitude, the steady state velocity decreases and steady state current
increases; however, the time constant does not change in all three cases.
6-1
6-4 The steady state speed is 4.716 rad/sec when the amplifier input voltage is 5V:
6-6
a) 6.25 rad/sec.
b) 63% of the steady state speed: 6.25 0.63 = 3.938 rad/sec
It takes 0.0249 seconds to reach 63% of its steady state speed.
c) The maximum current drawn by the motor is 1 Ampere.
6-1
6-7
a) 9.434 rad/sec.
b) 63% of the steady state speed: 9.434 0.63 = 5.943 rad/sec
It takes 0.00375 seconds to reach 63% of its steady state speed.
c) The maximum current drawn by the motor is 10 Amperes.
d) When there is no saturation, higher Kp value reduces the steady state error and decreases the rise
time. If there is saturation, the rise time does not decrease as much as it without saturation.
Also, if there is saturation and Kp value is too high, chattering phenomenon may appear.
6-8
a) The steady state becomes zero. The torque generated by the motor is 0.1 Nm.
b) 6.25 - (6.25 - 0) 0.63 = 2.31 rad/sec. It takes 0.0249 seconds to reach 63% of its new steady
state speed. It is the same time period to reach 63% of its steady state speed without the load
torque (compare with the answer for the Problem 6-6 b).
11-9 The SIMLab model becomes
The sensor gain and the speed input are reduced by a factor of 5. In order to get the same result as
Problem 6-6, the Kp value has to increase by a factor of 5. Therefore, Kp = 0.5.
The following graphs illustrate the speed and current when the input is 2 rad/sec and Kp = 0.5.
6-1
6-10
6-1
a) 1 radian.
b) 1.203 radians.
c) 0.2215 seconds.
6-11
6-1
a) The steady state position is very close to 1 radian.

b) 1.377 radians.
c) 0.148 seconds.
It has less steady state error and a faster rise time than Problem 6-10, but has larger overshoot.
6-1
6-12
Different proportional gains and their corresponding responses are shown on the following graph.
As the proportional gain gets higher, the motor has a faster response time and lower steady state error,
but if it the gain is too high, the motor overshoot increases. If the system requires that there be no
overshoot, Kp = 0.2 is the best value. If the system allows for overshoot, the best proportional gain is
dependant on how much overshoot the system can have. For instance, if the system allows for a 30%
overshoot, Kp = 1 is the best value.
6-1
6-13 Let Kp = 1 is the best value.
As the derivative gain increases, overshoot decreases, but rise time increases.
6-1
6-14
6-1
6-15
There could be many possible answers for this problem. One possible answer would be
Kp= 100
Ki= 10
Kd= 1.4
The Percent Overshoot in this case is 3.8%.
6-1
6-16
0.1 Hz
0.2 Hz
6-1
0.5 Hz
1 Hz
6-1
2Hz
5Hz
6-1
10Hz
50Hz
6-1
As frequency increases, the phase shift of the input and output also increase. Also, the amplitude of the
output starts to decrease when the frequency increases above 0.5Hz.
6-17
6-1
As proportional gain increases, the steady state error decreases.
6-1
6-18
Considering fast response time and low overshoot, Kp=1 is considered to be the best value.
6-1
6-19 It was found that the best Kp = 1
As Kd value increases, the overshoot decreases and the rise time increases.
6-1
6-20 From the Experiment Menu select the Open Loop Sine Input option.
6-1
Double click on the Sine wave block and choose the input values.
6-1
Run simulation in SIMULINK.

Change run time to 20 sec (default is
5 sec)
6-1
Calculate the Gain and Phase values by entering the input frequency. Repeat the process for other
frequency values and use the calculated gain and phase values to plot the frequency response of the
system.
6-1
6-21)
Select the Speed Control option in Virtual Lab Experiment Window.

Enter the Step Input and Controller Gain values by double clicking on their respective blocks.
6-1
Run simulation, and repeat the process for different gain values. Observe the steady state value change
with K.
6-1
6-1
6-22)
Select the Position Control option in Virtual Lab Experiment Window.

Enter the Step Input and Controller Gain values by double clicking on their respective blocks.
6-1
Run simulation, and repeat the process for different gain values.
6-1
6-1
6-23) If we choose Kp= 1, Ki= 0, and Kd= 0, we get
Increase Kd=0.1 we get:
6-1
6-1
6-24) Here Kp=275 and Kd=12.
Next Kp=275 and Kd=20.
6-1
You may try other parts and make observations.
6-1
6-25) a)
Change variables displayed using Setup Axes if desired.

b) Next use the Model Parameters button and change k to 15 as shown. Simulate the response and
show the desire variables.
6-1
6-1
6-1
c) A sample response
d)
6-1
Other parts are trivial and follow section 6-6.
6-1
6-1
Chapter 7
7-1 (a) P ( s ) = s 4 + 4 s 3 + 4 s 2 + 8 s Q( s ) = s + 1
Finite zeros of P(s): 0, 3.5098, 0.24512 j1.4897
Finite zeros of Q(s): 1
Asymptotes: K > 0: 60 o , 180 o , 300 o K < 0: 0 o , 120 o , 240 o
Intersect of Asymptotes:
3.5 0.24512 0.24512 ( 1)
1 = = 1
4 1
(b) P ( s ) = s 3 + 5 s 2 + s Q( s ) = s + 1
Finite zeros of P(s) : 0, 4.7912, 0.20871
Asymptotes: K > 0: 90 o , 270 o K < 0: 0 o , 180 o
4.7913 0.2087 ( 1)
1 = = 2
31
(c) P ( s ) = s 2 3 2
Q( s ) = s + 3s + 2 s + 8
Finite zeros of P(s): 0, 0
Finite zeros of Q(s): 3.156, 0.083156 j15874

.
Asymptotes: K > 0: 180 o K < 0: 0 o
71
(d) P ( s ) = s 3 + 2 s 2 + 3 s ( )
Q( s) = s 1 ( s + 3)
2
Finite zeros of P(s): 0, 1 j1.414
Finite zeros of Q(s): 1, 1, 3
Asymptotes: There are no asymptotes, since the number of zeros of P(s) and
Q(s) are
equal.
(e) P ( s ) = s 5 + 2 s 4 + 3 s 3 2
Q( s ) = s + 3s + 5
Finite zeros of P(s): 0, 0, 0, 1 j1.414

. j16583
.
1 1 ( 15)
. ( 15)
. 1
1 = =
5 2 3
(f) P ( s ) = s 4 + 2 s 2 + 10 Q( s ) = s + 5
Finite zeros of P(s): 1.0398 j1.4426, 10398

. j1.4426
10398
. 10398
. + 10398
. + 10398
. ( 5) 5
1 = =
4 1 3
72
7-2(a) MATLAB code:

s = tf('s')
num_GH=(s+1);
den_GH=(s^4+4*s^3+4*s^2+8*s);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
GH_p=pole(GH_a)
GH_z=zero(GH_a)
n=length(GH_p) %number of poles of G(s)H(s)
m=length(GH_z) %number of zeros of G(s)H(s)
%Asymptotes angles:
k=0;
Assymp1_angle=+180*(2*k+1)/(n-m)
Assymp2_angle=-180*(2*k+1)/(n-m)
k=1;
%Asymptotes intersection point on real axis:
sigma=(sum(GH_p)-sum(GH_z))/(n-m)
Root Locus
6
2
Imaginary Axis
-2
-4
-6
-8 -6 -4 -2 0 2 4
Real Axis
Assymp1_angle = 60
Assymp2_angle = -60
Assymp3_angle = 180
73
sigma = -1.0000 (intersect of asymptotes)

7-2(b) MATLAB code:
s = tf('s')
'Generating the transfer function:'
num_GH=(s+1);
den_GH=(s^3+5*s^2+s);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
GH_p=pole(GH_a)
GH_z=zero(GH_a)
%Asymptotes angles:
k=0;
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-5 -4 -3 -2 -1 0 1
Real Axis
Assymp1_angle = 90
Assymp2_angle = -90
74
sigma = -2 (intersect of asymptotes)
7-2(c) MATLAB code:

s = tf('s')
num_GH=(s^3+3*s^2+2*s+8);
den_GH=(s^2);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
GH_p=pole(GH_a)
GH_z=zero(GH_a)
%Asymptotes angles:
k=0;
Root Locus
2
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-6 -5 -4 -3 -2 -1 0 1
Real Axis
75
Assymp1_angle = 180
sigma = -3.0000 (intersect of asymptotes)
7-2(d) MATLAB code:

s = tf('s')
num_GH=((s^2-1)*(s+3));
den_GH=(s^3+2*s^2+3*s);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
GH_p=pole(GH_a)
GH_z=zero(GH_a)
Root Locus
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5
Real Axis
No asymptotes
76
7-2(e) MATLAB code:

s = tf('s')
num_GH=(s^2+3*s+5);
den_GH=(s^5+2*s^4+3*s^3);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
GH_p=pole(GH_a)
GH_z=zero(GH_a)
%Asymptotes angles:
k=0;
k=1;
Root Locus
6
2
Imaginary Axis
-2
-4
-6
-8 -6 -4 -2 0 2 4
Real Axis
Assymp1_angle = 60
77
Assymp2_angle = -60
Assymp3_angle = 180
sigma = 0.3333 (intersect of asymptotes)
7-2(f) MATLAB code:

s = tf('s')
num_GH=(s+5);
den_GH=(s^4+2*s^2+10);
GH_a=num_GH/den_GH;
figure(1);
rlocus(GH_a)
xlim([-20 20])
ylim([-20 20])
GH_p=pole(GH_a)
GH_z=zero(GH_a)
%Asymptotes angles:
k=0;
k=1;
78
Root Locus
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-20 -15 -10 -5 0 5 10 15 20
Real Axis
Assymp1_angle = 60
Assymp2_angle = -60
Assymp3_angle = 180
sigma = 1.6667 (intersect of asymptotes)
7-3) Consider
As the asymptotes are the behavior of G(s)H(s) when |s| , then
|s| > |zi| for i = 1,2,,m and |s| > |pi| for i = 1,2,,n
therefore G(s)H(s)
According to the condition on angles:
G(s)H(s) =
79
If we consider , then:
G(s)H(s) =
or
7-4) If , then each point on root locus must satisfy the characteristic equation of
If and , then
or
If the roots of above expression is considered as si for i = 1,2,,(n-m), then
since the intersect of (n-m) asymptotes lies on the real axis of the s-plane and is
real, therefore
7-5) Poles of GH is s = 0, -2, -1 + j, -1 j, therefore the center of asymptotes:
The angles of asymptotes:
710
j
j
711
7-6 (a) Angles of departure and arrival.
K > 0: 1 2 3 + 4 = 180o
1 90 45 + 90 = 180
o o o o
1 = 135
o
K < 0: 1 90o 45o + 90o = 0o
1 = 45
o
(b) Angles of departure and arrival.
1 2 3 + 4 = 180
o
K > 0:
1 135 90 + 90 = 0
o o o o
K < 0:
1 = 135
o
712
(c) Angle of departure:
1 2 3 + 4 = 180
o
K > 0:
1 135 90 45 = 180
o o o o
1 = 90
o
(d) Angle of departure
1 2 3 4 = 180
o
K > 0:
1 135 135 90 = 180

o o o o
1 = 180
o
713
(e) Angle of arrival
1 + 6 2 3 4 5 = 360
o
K < 0:
1 + 90 135 135 45 26.565 = 360

o o o o o o
1 = 108.435
o
714
7-7) a) G(s)H(s) =
we know that
therefore
As a result, D = G(s)H(s) 180o = 180 + , when -180o = 180o
b) Similarly:
G(s)H(s) =
Therefore:
As a result, =180
7-8) zeros: and poles:
Departure angles from:
715
Arrival angles at
7-9)(a) (b)
(c) (d)
716

717
718
7-10) The breaking points are on the real axis of and must satisfy
If and is a breakaway point, then
Finding where K is maximum or minimum , therefore
or
5 4 3 2
7-11) (a) Breakaway-point Equation: 2 s + 20 s + 74 s + 110 s + 48 s = 0
Breakaway Points: 0.7275, 2.3887
6 5 4 3 2
(b) Breakaway-point Equation: 3 s + 22 s + 65 s + 100 s + 86 s + 44 s + 12 = 0
Breakaway Points: 1, 2.5
6 5 4 3 2
(c) Breakaway-point Equation: 3 s + 54 s + 347.5 s + 925 s + 867.2 s 781.25 s 1953 = 0
719
Breakaway Points: 2.5, 109

.
6 5 4 3 2
(d) Breakaway-point Equation: s 8 s 19 s + 8 s + 94 s + 120 s + 48 = 0
Breakaway Points: 0.6428, 2.1208
7-12) (a)
K ( s + 8)
G( s) H ( s) =
s( s + 5)( s + 6)

o o o o
Asymptotes:K>0: 90 and 270 K<0: 0 and 180
IntersectofAsymptotes:
0 5 6 ( 8 )
1 = = 15
.
31
BreakawaypointEquation:
3 2
2 s + 35 s + 176 s + 240 = 0
BreakawayPoints: 2.2178, 5.5724, 9.7098
RootLocusDiagram:
720
712(b)
K
G( s) H ( s) =
s( s + 1)( s + 3)( s + 4 )
o o o o o o o o
Asymptotes:K>0: 45 , 135 , 225 , 315 K<0: 0 , 90 , 180 , 270
0 1 3 4
1 = = 2
4
3 2
BreakawaypointEquation: 4 s + 24 s + 38 s + 12 = 0
BreakawayPoints: 0.4189, 2, 3.5811
RootLocusDiagram:
721
712(c)
K( s + 4)
G( s) H ( s) = 2 2

s ( s + 2)
Asymptotes: K>0: 60 o , 180 o , 300 o K<0:

o o o
0 , 120 , 240
0 + 0 2 2 ( 4 )
1 = = 0
4 1
4 3 2
3 s + 24 s + 52 s + 32 s = 0
BreakawayPoints: 0, 1085
. , 2, 4.915
RootLocusDiagram:
722
712(d)
K ( s + 2)
G( s) H ( s) = 2

s( s + 2 s + 2 )
Asymptotes: K>0: 90 o , 270 o

o
K<0: 0 , 180
o
0 1 j 1 j ( 2 )
1 = = 0
31
3 2
BreakawaypointEquation: 2s + 8s + 8s + 4 = 0
BreakawayPoints: 2.8393 Theothertwosolutionsarenotbreakawaypoints.
723
RootLocusDiagram
712(e)
K ( s + 5)
G(s) H (s) =
(
s s + 2s + 2
2
)
Asymptotes: K>0: 90 o , 270 o K<0:
o o
0 , 180
0 1 j 1 j ( 5)
1 = = 15
.
31
3 2
2 s + 17 s + 20 s + 10 = 0
BreakawayPoints: 7.2091 Theothertwosolutionsarenotbreakawaypoints.
724
712(f)
K
G(s) H (s) =
(
s ( s + 4 ) s + 2s + 2
2
)
o o o o o o o o
Asymptotes:K>0: 45 , 135 , 225 , 315 K<0: 0 , 90 , 180 , 270
0 1 j 1+ j 4
1 = = 15
.
4
3 2
BreakawayPoint: 3.0922 Theothersolutonsarenotbreakawaypoints.
725
712(g)
2
K( s + 4)
G( s) H ( s) = 2 2

s ( s + 8)
Asymptotes: K>0: 90 o , 270 o o

K<0: 0 , 180
o
IntesectofAsymptotes:
0 + 0 8 8 ( 4 ) ( 4 )
1 =
42
5 4 3 2
BreakawaypointEquation: s + 20 s + 160 s + 640 s + 1040 s = 0
BreakawayPoints: 0,4,8,4j4,4+j4
726
712(h)
K
G( s) H ( s) = 2 2

s ( s + 8)
Asymptotes:K>0: 45o , 135o , 225o , 315o K<0: 0 o , 90 o , 180 o , 270 o
8 8
1 = = 4
4
3 2
BreakawaypointEquation: s + 12 s + 32 s = 0
BreakawayPoint: 0,4,8
727
712(i)
G(s) H (s) =
(
K s + 8s + 20
2
)
s ( s + 8)
2 2
Asymptotes:K>0: 90 o , 270 o K<0: 0 o , 180 o
8 8 ( 4 ) ( 4 )
1 = = 4
42
5 4 3 2
BreakawayPoints: 4, 8, 4 + j 4.9, 4 j 4.9
728
(j)
2
Ks
G(s) H (s) =
(s 2
4 )
Sincethenumberoffinitepolesandzerosof G ( s ) H ( s ) arethesame,therearenoasymptotes.
BreakawaypointEquation: 8 s = 0
BreakawayPoints: s=0
729
712(k)
G(s) H (s) =
K s 4 ( 2
)
(s 2
+1 )( s 2
+4 )
o o
0 , 180
2 + 2
IntersectofAsymptotes: 1 = = 0
42
6 4 2
BreakawaypointEquation: s 8 s 24 s = 0
BreakawayPoints: 0, 3.2132, 3.2132, j15246

. , j15246
.
730
712(l)
K ( s 1)
2
G(s) H (s) =
(s 2
+1 )( s 2
+4 )
o o o o
Asymptotes:K>0: 90 , 270 K<0: 0 , 180
1 + 1
1 = = 0
42
5 3
BreakawaypointEquation: s 2s 9s = 0
BreakawayPoints: 2.07, 2.07, j1.47, j1.47
731
(m)
K ( s + 1)( s + 2)( s + 3)
G( s) H ( s) = 3

s ( s 1)
732
o o
Asymptotes:K>0: 180 K<0: 0
6 5 4 3 2
BreakawaypointEquation: s + 12 s + 27 s + 2 s 18 s = 0
BreakawayPoints: 1.21,2.4,9.07,0.683,0,0
733
(n)
K ( s + 5)( s + 40)
G( s) H ( s) = 3

s ( s + 250)( s + 1000)
Asymptotes: K>0: 60 o , 180 o , 300 o K<0:

o o
0 , 120 , 240
o
Intersectofasymptotes:
0 + 0 + 0 250 1000 ( 5) ( 40 )
1 = = 40167
.
5 2
6 5 8 4 10 3 11 2
BreakawaypointEquation: 3750 s + 335000 s + 5.247 10 s + 2.9375 10 s + 1875
. 10 s = 0
BreakawayPoints: 7.288,712.2,0,0
734
712(o)
K ( s 1)
G( s) H ( s) =
s( s + 1)( s + 2 )

o o
0 , 180
735
1 2 1
1 = = 2
31
3
BreakawaypointEquation: s 3s 1 = 0
BreakawayPoints; 0.3473,1.532,1.879
736
Root Locus
7-13(a) MATLAB code: 10
num=[1 7]; 8
den=conv([1 0],[1 5]);
6
den=conv(den,[1 6]);
mysys=tf(num,den) 4
rlocus(mysys); 2
Imaginary Axis
0
-2
-4
-6
-8
-10
-8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
Root Locus
10
7-13(b) MATLAB code: 4
num=[0 1]; 2
Imaginary Axis
den=conv([1 0],[1 1]);

0
den=conv(den,[1 4]); -2
mysys=tf(num,den)
-4
rlocus(mysys);
-6
-8
-10
-12 -10 -8 -6 -4 -2 0 2 4 6 8
Real Axis
737
Root Locus
15
10
Imaginary Axis
7-13(c) MATLAB code: 0
num=[1 4];
-5
den=conv([1 0],[1 0]);
mysys=tf(num,den)
rlocus(mysys); -15
-12 -10 -8 -6 -4 -2 0 2 4 6 8
Real Axis
Root Locus
10
2
Imaginary Axis
-2
7-13(d) MATLAB code:
-4
num=[1 2];
-6
den=conv([1 0],[1
(1+j)]); -8
den=conv(den,[1 (1-
-10
j)]); -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
mysys=tf(num,den)
rlocus(mysys);
738
Root Locus
25
20
15
10
5
Imaginary Axis
0
-5
-10
-15
7-13(e) MATLAB code: -20
num=[1 5]; -25

-6 -5 -4 -3 -2 -1 0 1 2
den=conv([1 0],[1 Real Axis
(1+j)]);
den=conv(den,[1 (1-
j)]);
mysys=tf(num,den)
rlocus(mysys);
739
Root Locus
5
7-13(f) MATLAB code: 4
num=conv([1 4],[1 4]); 3

den=conv([1 0],[1 0]);
2
den=conv(den,[1 8]); 1
Imaginary Axis
mysys=tf(num,den) 0
rlocus(mysys);
-1
-2
-3
-4
-5
-6 -5 -4 -3 -2 -1 0 1 2
Real Axis
Root Locus
10
8
7-13(g) MATLAB code:
6
num=conv([1 4],[1 4]);
den=conv([1 0],[1 0]); 4
Imaginary Axis
0
mysys=tf(num,den)
rlocus(mysys); -2
-4
-6
-8
-10
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
740
Root Locus
25
20
15
7-13(h) MATLAB code: 10
num=[0 1]; 5
Imaginary Axis
den=conv([1 0],[1 0]);
mysys=tf(num,den)
-10
rlocus(mysys);
-15
-20
-25
-30 -25 -20 -15 -10 -5 0 5 10 15 20
Real Axis
Root Locus
15
10
5
Imaginary Axis
7-13(i) MATLAB code:

0
num=conv([1 4-2j],[1
4+2j])
-5
den=conv([1 0],[1 0]);
mysys=tf(num,den)
rlocus(mysys);
-15
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
741
Root Locus
0.25
0.2
0.15
0.1
0.05
Imaginary Axis 0
7-13(j) MATLAB code: -0.05
num=conv([1 0],[1 0]); -0.1

den=conv([1 2],[1 -2]);
-0.15
mysys=tf(num,den)
rlocus(mysys); -0.2
-0.25
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Real Axis
7-13(k) MATLAB code:

num=conv([1 2],[1 -2]);
den=conv([1 -j],[1 j]);
den=conv(den,[1 -2j]);
den=conv(den,[1 2j]);
mysys=tf(num,den)
742
rlocus(mysys);
Root Locus
5
Imaginary Axis
0
-1
-2
-3
-4
-5
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Real Axis
7-13(l) MATLAB code:

num=conv([1 1],[1 -1]); Root Locus
5
den=conv([1 -j],[1 j]);
den=conv(den,[1 -2j]); 4
den=conv(den,[1 2j]); 3
mysys=tf(num,den)
rlocus(mysys); 2
1
Imaginary Axis
-1
-2
-3
-4
-5
-1.5 -1 -0.5 0 0.5 1 1.5
Real Axis
743
Root Locus
7-13(m) MATLAB code: 6
num=conv([1 1],[1 2]);

4
num=conv(num,[1 3]);
den=conv([1 0],[1 0]);
den=conv(den,[1 -1]);
Imaginary Axis
mysys=tf(num,den) 0
rlocus(mysys);
-2
-4
-6
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
Root Locus
1500
1000
7-13(n) MATLAB code:
num=conv([1 5],[1 40]); 500
den=conv([1 0],[1 0]);
Imaginary Axis
0
mysys=tf(num,den) -500
rlocus(mysys);
-1000
-1500
-1400 -1200 -1000 -800 -600 -400 -200 0 200 400 600
Real Axis
744
Root Locus
10
Imaginary Axis
7-13(o) MATLAB code: 0
num=conv([1 5],[1 40]); -2

den=conv([1 0],[1 0]);
-4
-8
mysys=tf(num,den)
rlocus(mysys); -10
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5
Real Axis
7-14)(a) Q( s ) = s + 5 ( )
P ( s ) = s s + 3s + 2 = s ( s + 1)( s + 2)
2

o
0 , 180
o
1 2 ( 5)
1 = = 1
31
3 2
BreakawaypointEquation: s + 9 s + 15 s + 5 = 0
BreakawayPoints: 0.4475,1.609,6.9434
745
714(b) Q( s ) = s + 3 2

P( s) = s s + s + 2
746
o o o o
Asymptotes: K>0: 90 , 270 K<0: 0 , 180
1 ( 3)
1 = = 1
31
3 2
BreakawaypointEquation: s + 5s + 3s + 3 = 0
BreakawayPoints: 4.4798 Theothersolutionsarenotbreakawaypoints.
747
2
714(c) Q( s ) = 5s P ( s ) = s + 10
o o
Asymptotes: K>0: 180 K<0: 0
748
2
BreakawaypointEquation: 5 s 50 = 0
BreakawayPoints: 3.162, 3.162
749
714(d) (
Q( s ) = s s + s + 2
2
) P ( s ) = s + 3s + s + 5s + 10
4 3 2
o o
Asymptotes:K>0: 180 K<0: 0
6 5 4 3 2
BreakawaypointEquation: s + 2 s + 8 s + 2 s 33 s 20 s 20 = 0
BreakawayPoints: 2,1.784. Theothersolutionsarenotbreakawaypoints.
750
714(e) ( )
Q( s ) = s 1 ( s + 2 )
2
(
P( s) = s s + 2s + 2
2
)
751
SinceQ( s ) and P ( s ) areofthesameorder,therearenoasymptotes.
3 2
BreakawayPoints: 1.3848
752
714(f) Q ( s ) = ( s + 1)( s + 4) ( )
P(s) = s s 2
2
753
o o
Asymptotes: K>0: 180 K<0: 0
4 3 2
Breakawaypointequations: s + 10 s + 14 s 8 = 0
BreakawayPoints: 8.334, 0.623
754
714(g) Q( s ) = s + 4 s + 5
2
P( s ) = s
2
(s 2
)
+ 8s + 16
Asymptotes: K>0: 90 o , 270 o o o

K<0: 0 , 180
8 ( 4 )
1 = = 2
42
5 4 3 2
BreakawayPoints: 0, 2, 4, 2 + j 2.45, 2 j 2.45
755
714(h) (
Q( s ) = s 2
2
) ( s + 4) (
P( s) = s s + 2s + 2
2
)
Since Q ( s ) and P ( s ) areofthesameorder,therearenoasymptotes.
756
BreakawayPoints: 2,6.95
757
714(i) Q( s ) = ( s + 2 )( s + 0.5)
2
P( s) = s s 1
Asymptotes: K>0: 180 o K<0:

o
0
4 3 2
BreakawaypointEquation: s + 5s + 4 s 1 = 0
BreakawayPoints: 4.0205,0.40245 Theothersolutionsarenotbreakawaypoints.
758
759
714(j) Q( s) = 2s + 5 P(s) = s
2
(s 2
)
+ 2s + 1 = s
2
( s + 1)
2
Asymptotes: K>0: 60 o , 180 o , 300 o o o

K<0: 0 , 120 , 240
o
IntersectofAsymptotes;
0 + 0 1 1 ( 2.5) 0.5
1 = = = 0.167
4 1 3
4 3 2
BreakawaypointEquation: 6 s + 28 s + 32 s + 10 s = 0
BreakawayPoints: 0,0.5316,1,3.135
760
7-15 ) MATLAB code:

clear all;
close all;
s = tf('s')
%a)
num_GH_a=(s+5);
den_GH_a=(s^3+3*s^2+2*s);
GH_a=num_GH_a/den_GH_a;
figure(1);
rlocus(GH_a)
761
%b)
num_GH_b=(s+3);
den_GH_b=(s^3+s^2+2*s);
GH_b=num_GH_b/den_GH_b;
figure(2);
rlocus(GH_b)
%c)
num_GH_c= 5*s^2;
den_GH_c=(s^3+10);
GH_c=num_GH_c/den_GH_c;
figure(3);
rlocus(GH_c)
%d)
num_GH_d=(s^3+s^2+2);
den_GH_d=(s^4+3*s^3+s^2+15);
GH_d=num_GH_d/den_GH_d;
figure(4);
rlocus(GH_d)
%e)
num_GH_e=(s^2-1)*(s+2);
den_GH_e=(s^3+2*s^2+2*s);
GH_e=num_GH_e/den_GH_e;
figure(5);
rlocus(GH_e)
%f)
num_GH_f=(s+4)*(s+1);
den_GH_f=(s^3-2*s);
GH_f=num_GH_f/den_GH_f;
figure(6);
rlocus(GH_f)
%g)
num_GH_g=(s^2+4*s+5);
den_GH_g=(s^4+6*s^3+9*s^2);
GH_g=num_GH_g/den_GH_g;
figure(7);
rlocus(GH_g)
%h)
num_GH_h=(s^2-2)*(s+4);
den_GH_h=(s^3+2*s^2+2*s);
GH_h=num_GH_h/den_GH_h;
762
figure(8);
rlocus(GH_h)
%i)
num_GH_i=(s+2)*(s+0.5);
den_GH_i=(s^3-s);
GH_i=num_GH_i/den_GH_i;
figure(9);
rlocus(GH_i)
%j)
num_GH_j=(2*s+5);
den_GH_j=(s^4+2*s^3+2*s^2);
GH_j=num_GH_j/den_GH_j;
figure(10);
rlocus(GH_j)
%k)
num_GH_k=1;
den_GH_k=(s^5+2*s^4+3*s^3+2*s^2+s);
GH_k=num_GH_k/den_GH_k;
figure(11);
rlocus(GH_k)
Root Locus diagram 7-15(a):
Root Locus
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-6 -5 -4 -3 -2 -1 0 1
Real Axis
763
Root Locus diagram 7-15(b):
Root Locus
15
10
5
Imaginary Axis
-5
-10
-15
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1
Real Axis
Root Locus diagram 7-15(c):
764
Root Locus
2
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-5 -4 -3 -2 -1 0 1 2
Real Axis
Root Locus diagram 7-15(d):
Root Locus
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-6 -5 -4 -3 -2 -1 0 1 2
Real Axis
Root Locus diagram 7-15(e):
765
Root Locus
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5
Real Axis
Root Locus diagram 7-15(f):

poles:
Asymptotes angle:
Therefore,
Departure angle from:
766
Root Locus
5
1
Imaginary Axis
-1
-2
-3
-4
-5
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
Root Locus diagram 7-15(g):

Poles: and zeroes:
Asymptotes angles:
767
Root Locus
15
10
5
Imaginary Axis
-5
-10
-15
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
Root Locus diagram 7-15(h):

Poles: and zeros:
The break away points:
which means:
or
768
Root Locus
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-5 -4 -3 -2 -1 0 1 2
Real Axis
Root Locus diagram 7-15(i):

Poles:
breaking points:
which means :
Asymptotes angles:
or =
769
Root Locus
2.5
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-2.5
-10 -8 -6 -4 -2 0 2
Real Axis
Root Locus diagram 7-15(j):
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-10 -5 0 5
Real Axis
Root Locus diagram 7-15(k):
770
Root Locus
6
2
Imaginary Axis
-2
-4
-6
-6 -5 -4 -3 -2 -1 0 1 2 3 4
Real Axis
7-16) (a)Asymptotes: K>0: 45o , 135o , 225o , 315o
2 2 5 6 ( 4 )
1 = = 2.75
51
5 4 3 2
BreakawaypointEquation: 4 s + 65 s + 396 s + 1100 s + 1312 s + 480 = 0
BreakawayPoints: 0.6325,5.511(ontheRL)
When = 0.707, K=13.07
771
716(b)Asymptotes: K>0: 45o , 135o , 225o , 315o
772
0 2 5 10
1 = = 4.25
4
3 2
BreakawaypointEquation: 4 s + 51s + 160 s + 100 = 0
When = 0.707, K=61.5
773
716(c)Asymptotes: K>0: 180 o
4 3 2
BreakawaypointEquation: s + 4 s + 10 s + 300 s + 500 = 0
BreakawayPoints: 1.727 (ontheRL)
When = 0.707, K=9.65
774
775
o o
716(d)K>0: 90 , 270
2 2 5 6
1 = = 7.5
42
When = 0.707, K=8.4
776
7-17) MATLAB code:

clear all;
close all;
s = tf('s')
%a)
num_G_a=(s+3);
den_G_a=s*(s^2+4*s+4)*(s+5)*(s+6);
G_a=num_G_a/den_G_a;
figure(1);
rlocus(G_a)
%b)
777
num_G_b= 1;
den_G_b=s*(s+2)*(s+4)*(s+10);
G_b=num_G_b/den_G_b;
figure(2);
rlocus(G_b)
%c)
num_G_c=(s^2+2*s+8);
den_G_c=s*(s+5)*(s+10);
G_c=num_G_c/den_G_c;
figure(3);
rlocus(G_c)
%d)
num_G_d=(s^2+4);
den_G_d=(s+2)^2*(s+5)*(s+6);
G_d=num_G_d/den_G_d;
figure(4);
rlocus(G_d)
%e)
num_G_e=(s+10);
den_G_e=s^2*(s+2.5)*(s^2+2*s+2);
G_e=num_G_e/den_G_e;
figure(5);
rlocus(G_e)
%f)
num_G_f=1;
den_G_f=(s+1)*(s^2+4*s+5);
G_f=num_G_f/den_G_f;
figure(6);
rlocus(G_f)
%g)
num_G_g=(s+2);
den_G_g=(s+1)*(s^2+6*s+10);
G_g=num_G_g/den_G_g;
figure(7);
rlocus(G_g)
%h)
num_G_h=(s+3)*(s+2);
den_G_h=s*(s+1);
G_h=num_G_h/den_G_h;
figure(8);
rlocus(G_h)
778
%i)
num_G_i=1;
den_G_i=s*(s^2+4*s+5);
G_i=num_G_i/den_G_i;
figure(9);
rlocus(G_i)

By using Data Cursor tab on the figure window and clicking on the root locus diagram, gain and
damping values can be observed. Damping of ~0.707 can be observed on intersection of the root locus
diagram with two lines originating from (0,0) by angles of ArcCos(0.707) from the real axis. These
intersection points are shown for part (a) where the corresponding gain is 19. In the other figures for
section (b) to (i), similar points have been picked by the Data Cursor, and the gains are reported here.
Root Locus
15
10
System: G_a
Gain: 19
Pole: -0.584 + 0.589i
5 Damping: 0.704
Overshoot (%): 4.43
Imaginary Axis
Frequency (rad/sec): 0.829
System: G_a
Gain: 19
-5 Pole: -0.584 - 0.589i
Damping: 0.704
Overshoot (%): 4.43
-10
-15
-20 -15 -10 -5 0 5 10 15
Real Axis
Root Locus diagram 7-17(b): (K = 45.5 @ damping = ~0.0707)
779
Root Locus
25
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-25
-30 -25 -20 -15 -10 -5 0 5 10 15 20
Real Axis
Root Locus diagram 7-17(c): (K = 12.8 @ damping = ~0.0707)
Root Locus
3
1
Imaginary Axis
-1
-2
-3
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
Root Locus diagram 7-17(d): (K = 8.3 @ damping = ~0.0707)
780
Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
Root Locus diagram 7-17(e): (K = 0 @ damping = 0.0707)
Root Locus
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-15 -10 -5 0 5 10 15 20
Real Axis
Root Locus diagram 7-17(f): (K = 2.33 @ damping = ~0.0707)
781
Root Locus
5
1
Imaginary Axis
-1
-2
-3
-4
-5
-7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
Root Locus diagram 7-17(g): (K = 7.03 @ damping = ~0.0707)
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
Root Locus diagram 7-17(h): (no solution exists for damping =0.0707)
782
Root Locus
1
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1
-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
Root Locus diagram 7-17(i): (K = 2.93 @ damping = ~0.0707)
Root Locus
4
1
Imaginary Axis
-1
-2
-3
-4
-6 -5 -4 -3 -2 -1 0 1
Real Axis
o o o
7-18) (a)Asymptotes: K>0: 60 , 180 , 300
783
0 10 20
1 = = 10
3
2
BreakawaypointEquation: 3 s + 60 s + 200 = 0 BreakawayPoint:(RL)4.2265,K=384.9

o o o o
(b)Asymptotes: K>0: 45 , 135 , 225 , 315
0 1 3 5
1 = = 2.25
4
3 2
BreakawayPoints:(RL)0.4258 K=2.879, 4.2537K=12.95
784
c) Zeros: and poles:
Angle of asymptotes:
The breakaway points:
Then and
d) Poles:
breakaway points:
e) Zeros: and poles:
785
breakaway points:
f) Poles:
Angles of asymptotes:
breakaway point:
7-19) MATLAB code:

clear all;
close all;
s = tf('s')
%a)
num_G_a=1;
den_G_a=s*(s+10)*(s+20);
figure(1);
rlocus(G_a)
%b)
num_G_b= 1;
den_G_b=s*(s+1)*(s+3)*(s+5);
figure(2);
rlocus(G_b)
%c)
num_G_c=(s-0.5);
den_G_c=(s-1)^2;
figure(3);
rlocus(G_c)
%d)
786
num_G_d=1;
den_G_d=(s+0.5)*(s-1.5);
figure(4);
rlocus(G_d)
%e)
num_G_e=(s+1/3)*(s+1);
den_G_e=s*(s+1/2)*(s-1);
figure(5);
rlocus(G_e)
%f)
num_G_f=1;
den_G_f=s*(s^2+6*s+25);
G_f=num_G_f/den_G_f;
figure(6);
rlocus(G_f)
Root Locus
60
40
System: G_a
Gain: 385
20 Pole: -4.23 - 1.45e-007i
Damping: 1
Overshoot (%): 0
Imaginary Axis

0
-20
-40
-60
-80 -60 -40 -20 0 20 40
Real Axis
787
Root Locus
15
10
System: G_b
Gain: 2.88
5 Pole: -0.426 - 1.02e-008i
Damping: 1
Overshoot (%): 0
Imaginary Axis

0
System: G_b
Gain: 13
Pole: -4.25 + 0.0278i
-5 Damping: 1
Overshoot (%): 0
-10
-15
-15 -10 -5 0 5 10
Real Axis
Root Locus
0.8
0.6
System: G_c
0.4
Gain: 2
Pole: 0
0.2 Damping: -1
Overshoot (%): 0
Imaginary Axis
Frequency (rad/sec): 0
0
-0.2
-0.4
-0.6
-0.8
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2
Real Axis
788
Root Locus
1.5
1
System: G_d
Gain: 1
0.5 Pole: 0.5
Damping: -1
Overshoot (%): 0
Imaginary Axis

0
-0.5
-1
-1.5
-1 -0.5 0 0.5 1 1.5 2
Real Axis
Root Locus
1.5
1
System: G_e System: G_e
Gain: 5.34 Gain: 0.211
0.5 Pole: -2.24 - 3.95e-008i Pole: 0.383
Damping: 1 Damping: -1
Overshoot (%): 0 Overshoot (%): 0
Imaginary Axis
Frequency (rad/sec): 2.24 Frequency (rad/sec): 0.383

0
-0.5
-1
-1.5
-5 -4 -3 -2 -1 0 1 2
Real Axis
Root Locus diagram 7-19(f): (No breakaway points)
789
Root Locus
25
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-25
-30 -25 -20 -15 -10 -5 0 5 10 15
Real Axis
790
7-20)
791
7-21) MATLAB code:

clear all;
close all;
s = tf('s')
%a)
n=1;
num_G_a= 1;
den_G_a=(s+4)^n;
figure(n);
rlocus(G_a)
%b)
n=2;
num_G_b= 1;
den_G_b=(s+4)^n;
figure(n);
rlocus(G_b)
%c)
n=3;
num_G_c= 1;
den_G_c=(s+4)^n;
figure(n);
rlocus(G_c)
%d)
n=4;
num_G_d= 1;
den_G_d=(s+4)^n;
figure(n);
rlocus(G_d)
%e)
n=5;
num_G_e= 1;
den_G_e=(s+4)^n;
figure(n);
rlocus(G_e)
792
Root Locus
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-16 -14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
Root Locus
2.5
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-2
-2.5
-4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
793
Root Locus
4
1
Imaginary Axis
-1
-2
-3
-4
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
Root Locus
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-6 -5 -4 -3 -2 -1 0 1
Real Axis
794
Root Locus
5
1
Imaginary Axis
-1
-2
-3
-4
-5
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1
Real Axis
795
7-22) P ( s ) = s 3 + 25 s 2 + 2 s + 100 Q ( s ) = 100 s
o o
Asymptotes: Kt > 0: 90 , 270
25 0
1 = = 12.5
31
3 2
BreakawaypointEquation: s + 12.5 s 50 = 0
BreakawayPoints: (RL) 2.2037, 12.162
7-23) MATLAB code:
796
s = tf('s')
num_G= 100;
den_G=s^3+25*s+2*s+100;
G=num_G/den_G;
figure(1);
rlocus(G)
Root Locus diagram 7-23:
Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-30 -25 -20 -15 -10 -5 0 5 10 15 20
Real Axis
7-24)Characteristicequation: s 3 + 5s 2 + Kt s + K = 0
(a) K t = 0 : P(s) = s
2
( s + 5) Q(s) = 1
o o o
Asymptotes:K>0: 60 , 180 , 300
5 0
1 = = 1667
.
3
797
2
BreakawaypointEquation: 3 s + 10 s = 0 BreakawayPoints: 0,3.333
798
3 2
724(b) P ( s ) = s + 5 s + 10 = 0 Q( s ) = s
Asymptotes:K>0: 90 o , 270 o
5 0
1 = = 0
2 1
3
BreakawaypointEquation: 2 s + 5 s 10 = 0
TherearenobreakawaypointsonRL.
799
7-25)
By collapsing the two loops, and finding the overall close loop transfer function, the characteristic
equation (denominator of closed loop transfer function) can be found as:
s 3 + 5s 2 + K t s + K
1 + GH =
s 2 ( s + 5) + K t s
For part (a): Root locus diagram, part (a):

Kt =0. Therefore, assuming 15
Root Locus
Den(GH)= s 3 + 5s 2 and
10
Num(GH) =1, we can use rlocus
command to construct the root 5
locus diagram.
Imaginary Axis
For part (b): -5
K =10. Therefore, assuming

-10
Den(GH)= s 3 + 5 s 2 + 10 and
Num(GH) = s , we can use rlocus -15
-20 -15 -10 -5 0 5 10
command to construct the root Real Axis
locus diagram. Root locus diagram, part (b):

Root Locus
15
MATLAB code (7-25): 10
s = tf('s')
5
%a)
Imaginary Axis
num_G_a= 1;
den_G_a=s^3+5*s^2; 0
GH_a=num_G_a/den_G_a;
figure(1); -5
rlocus(GH_a)
-10
%b)
num_G_b= s;
den_G_b=s^3+5*s^2+10; -15
-6 -5 -4 -3 -2 -1 0 1
GH_b=num_G_b/den_G_b; Real Axis
figure(2);
rlocus(GH_b)
7100
7-26) P ( s ) = s 2 + 116.84 s + 1843 2

Q ( s ) = 2.05 s ( s + 5)
o
Asymptotes: J L = 0: 180
4 3 2
BreakawaypointEquation: 2.05 s 479 s 12532 s 37782 s = 0
BreakawayPoints: (RL) 0,204.18
7-27) MATLAB code:
7101
s = tf('s')
num_G = (2.05*s^3 + 10.25*s^2);
den_G = (s^2 + 116.84*s + 1843);
G = num_G/den_G;
figure(1);
rlocus(G)
Root locus diagram:
Root Locus
150
100
50
Imaginary Axis
-50
-100
-150
-400 -350 -300 -250 -200 -150 -100 -50 0 50
Real Axis
7102
7-28) (a) P( s ) = s ( s 2 1) Q ( s ) = ( s + 5)( s + 3)
o
Asymptotes: K>0: 180
4 3 2
BreakawaypointEquation: s + 16 s + 46 s 15 = 0
BreakawayPoints: (RL) 0.5239,12.254
7103
728(b) P( s ) = s ( s 2 + 10 s + 29 ) Q ( s ) = 10( s + 3)
Asymptotes: K>0: 90 o , 270 o
0 10 ( 3)
1 = = 3.5
31
3 2
TherearenobreakawaypointsontheRL.
7104
7-29)
Root locus diagram, part (a):
MATLAB code (7-29):
s = tf('s')
%a)
num_G_a = (s+5)*(s+3);
den_G_a = s*(s^2 - 1);
G_a = num_G_a/den_G_a;
figure(1);
rlocus(G_a)
7105
K=10;
%b) Root Locus
8
num_G_b = (3*K+K*s);
den_G_b =
6
(s^3+K*s^2+K*3*s-s);
G_b = num_G_b/den_G_b; 4
figure(2);
rlocus(G_b) 2
Imaginary Axis
0
-2
-4
-6
-8
-25 -20 -15 -10 -5 0 5
Real Axis
Root locus diagram, part (b):
Root Locus
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-6 -5 -4 -3 -2 -1 0 1
Real Axis
7-30) Poles: zeros:
7106
Angles of asymptotes:
breakaway points:
MATLAB code:
s = tf('s')
num_G=(s+0.4);
den_G=s^2*(s+3.6);
G=num_G/den_G;
figure(1);
rlocus(G)
Rootlocusdiagram:
Root Locus
5
1
Imaginary Axis
-1
-2
-3
-4
-5
-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
731(a) P ( s ) = s( s + 12.5)( s + 1) Q( s ) = 83.333
7107
o o o
Asymptotes: N>0: 60 , 180 , 300
0 12.5 1
1 = = 4.5
3
2
BreakawaypointEquation: 3s + 27 s _12.5 = 0
BreakawayPoint:(RL) 0.4896

2 2
731(b) P ( s ) = s + 12.5 s + 833.333 Q ( s ) = 0.02 s ( s + 12.5)
7108
o
A>0:180
4 3 2
BreakawaypointEquation: 0.02 s + 0.5 s + 53.125 s + 416.67 s = 0
BreakawayPoints:(RL)0
731c) P ( s ) = s 3 + 12.5 s 2 + 1666.67 = ( s + 17.78 )( s 2.64 + j 9.3)( s 2.64 j 9.3)
7109
Q( s ) = 0.02 s( s + 12.5)
o
Asymptotes: Ko > 0: 180
4 3 2
BreakawaypointEquation: 0.02 s + 0.5 s + 3.125 s 66.67 s 416.67 = 0
BreakawayPoint:(RL) 5.797
7110
7-32) MATLAB code:
s = tf('s')
%a)
A=50;
K0=50;
num_G_a = 250;
den_G_a = 0.06*s*(s + 12.5)*(A*s+K0);
figure(1);
rlocus(G_a)
%b)
N=10;
K0=50;
num_G_b = 0.06*s*(s+12.5)*s
den_G_b = K0*(0.06*s*(s+12.5))+250*N;
G_b = num_G_b/den_G_b;
figure(2);
rlocus(G_b)
%c)
A=50;
N=20;
num_G_c = 0.06*s*(s+12.5);
den_G_c = 0.06*s*(s+12.5)*A*s+250*N;
G_c = num_G_c/den_G_c;
figure(3);
rlocus(G_c)
7111
Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-40 -30 -20 -10 0 10 20
Real Axis
Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-25 -20 -15 -10 -5 0 5
Real Axis
Root locus diagram, part (c):
7112
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-35 -30 -25 -20 -15 -10 -5 0 5
Real Axis
7-33) (a) A = Ko = 100: P ( s ) = s( s + 12.5)( s + 1) Q( s ) = 4167

.
o o o
Asymptotes: N>0: 60 180 300
0 1 12.5
1 = = 4.5
3
2
BreakawaypointEquation: 3 s + 27 s + 12.5 = 0
BreakawayPoints:(RL) 0.4896
7113
733(b) P ( s ) = s 2 + 12.5 + 1666.67 = ( s + 6.25 + j 40.34 )( s + 6.25 j 40.34 )
Q ( s ) = 0.02 s 2 ( s + 12.5)
7114
Asymptotes:A>0:180 o
4 3 2
BreakawaypointEquation: 0.02 s + 0.5 s + 103.13 s + 833.33 s = 0
BreakawayPoints: (RL) 0
7115
733(c) P ( s ) = s 3 + 12.5 s 2 + 833.33 = ( s + 15.83)( s 1.663 + j 7.063)( s 1.663 j 7.063)
Q( s ) = 0.01s( s + 12.5)
Asymptotes: Ko > 0: 180o
4 3 2
BreakawaypointEquation: 0.01s + 0.15 s + 1.5625 s 16.67 s 104.17 = 0
BreakawayPoint: (RL) 5.37
7116
7-34) MATLAB code:
s = tf('s')
%a)
A=100;
K0=100;
num_G_a = 250;
den_G_a = 0.06*s*(s + 12.5)*(A*s+K0);
figure(1);
rlocus(G_a)
%b)
N=20;
K0=50;
7117
num_G_b = 0.06*s*(s+12.5)*s
den_G_b = K0*(0.06*s*(s+12.5))+250*N;
G_b = num_G_b/den_G_b;
figure(2);
rlocus(G_b)
%c)
A=100;
N=20;
num_G_c = 0.06*s*(s+12.5);
den_G_c = 0.06*s*(s+12.5)*A*s+250*N;
G_c = num_G_c/den_G_c;
figure(3);
rlocus(G_c)

Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-40 -30 -20 -10 0 10 20
Real Axis
7118
Root Locus
50
40
30
20
10
Imaginary Axis
-10
-20
-30
-40
-50
-30 -25 -20 -15 -10 -5 0 5 10
Real Axis
Root locus diagram, part (c):
Root Locus
8
2
Imaginary Axis
-2
-4
-6
-8
-30 -25 -20 -15 -10 -5 0 5
Real Axis
7-35) a) zeros: , poles:
7119
Breakaway points:
b) There is no closed loop pole in the right half s-plane; therefore the system is stable for all K>0
c) MATLAB code:
num_G=25*(s+2)^2;
den_G=(s^2+4)*(s+5)^2;
G_a=num_G/den_G;
figure(1);
rlocus(G_a)
Root locus diagram:

Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-6 -5 -4 -3 -2 -1 0 1
Real Axis
7120
2
7-36)(a) P ( s ) = s ( s + 1)( s + 5) Q ( s ) = 1
o o o o
Asymptotes:K>0: 45 , 135 , 225 , 315
0 + 0 1 5
1 = = 15
.
4
3 2
BreakawaypointEquation: 4 s + 18 s + 10 s = 0 Breakawaypoint:(RL)0,3.851
2
(b) P ( s ) = s ( s + 1)( s + 5) Q ( s ) = 5 s + 1
o o o
Asymptotes: K>0: 60 , 180 , 300
0 + 0 1 5 ( 0.2) 5.8
1 = = = 193
.
4 1 3
4 3 2
BreakawaypointEquation: 15 s + 64 s + 43 s + 10 s = 0
BreakawayPoints:(RL) 3.5026
7121
7-37) Root locus diagram, part (a):

MATLAB code (7-37):
Root Locus
s = tf('s') 15
%a)
num_GH_a= 1;
10
den_GH_a=s^2*(s+1)*(s+
5);
GH_a=num_GH_a/den_GH_a 5
;
Imaginary Axis
figure(1); 0
rlocus(GH_a)
-5
%b)
num_GH_b= (5*s+1);
den_GH_b=s^2*(s+1)*(s+ -10
5);
GH_b=num_GH_b/den_GH_b
-15
; -15 -10 -5 0 5 10 15
figure(2); Real Axis
rlocus(GH_b)
7122
Root Locus
20
15
10
Imaginary Axis
0
-5
-10
-15
-20
-25 -20 -15 -10 -5 0 5 10
Real Axis
7-38) a) can be approximated by ( easy way to verify is to compare both funtions Taylor
series expansions)
Therefore:
Zeros: and poles:

Angle of asymptotes :
Breakaway points:
Which means:
b)
S2 1 2+2k
7123
S 3-k 0
S0 (3-k)(2+2k)

As a result:
Since K must be positive, the range of stability is then
c) In this problem, e Ts term is a time delay. Therefore, MATLB PADE command is used for pade
approximation, where brings e Ts term to the polynomial form of degree N.
s = tf('s')
T=1
N=1;
num_GH= pade(exp(-1*T*s),N);
den_GH=(s+1);
GH=num_GH/den_GH;
figure(5);
rlocus(GH)
Root locus diagram:
7124
Root Locus
4
1
Imaginary Axis
-1
-2
-3
-4
-4 -2 0 2 4 6 8 10 12
Real Axis
7-39)
(a) P ( s ) = s 2 ( s + 1)( s + 5) + 10 = ( s + 4.893)( s + 1896
. )( s 0.394 + j 0.96 )( s 0.394 + j 0.96 )
Q( s ) = 10 s
o o o
Asymptotes: Td > 0: 60 , 180 , 300
4.893 1896
. + 0.3944 + 0.3944
IntersectionofAsymptotes: 1 = = 2
4 1
TherearenobreakawaypointsontheRL.
7125
(b) MATLAB code:

s = tf('s')
num_GH= 10*s;
den_GH=s^2*(s+1)*(s+5)+10;
GH=num_GH/den_GH;
figure(1);
rlocus(GH)
Root locus diagram:
7126
Root Locus
15
10
5
Imaginary Axis
-5
-10
-15
-20 -15 -10 -5 0 5 10
Real Axis
7-40) (a)K=1: P ( s ) = s 3 ( s + 117.23)( s + 4882.8 ) Q ( s ) = 1010( s + 1.5948 )( s + 114.41)( s + 4884 )
o o
Asymptotes: K L > 0: 90 , 270
117.23 4882.8 + 15948

. + 114.41 + 4884
1 = = 0.126
5 3
BreakawayPoint:(RL)0
7127
3
740(b)K=1000: P ( s ) = s ( s + 117.23)( s + 4882.8 )
3 2 5
Q ( s ) = 1010( s + 5000 s + 5.6673 10 s + 891089110 )

= 1010( s + 4921.6 )( s + 39.18 + j 423.7 )( s + 39.18 423.7 )
o o
Asymptotes: K L > 0: 90 , 270
117.23 4882.8 + 49216

. + 39.18 + 39.18
1 = = 0.033
5 3
7 7 6 10 5 13 4 16 3 18 2
2020 s + 2.02 10 s + 5.279 10 s + 1.5977 10 s + 18655
. 10 s + 1.54455 10 s = 0
Breakawaypoints:(RL) 0,87.576
7128
7-41) MATLAB code:

s = tf('s')
Ki=9;
Kb=0.636;
Ra=5;
La=.001;
Ks=1;
n=.1;
Jm=0.001;
Jl=0.001;
Bm=0;
%a)
7129
K=1;
num_G_a=((n^2*La*Jl+La*Jm)*s^3+(n^2*Ra*Jl+Ra*Jm+Bm*La)*s^2+Ra*Bm*s+Ki
*Kb*s+n*Ks*K*Ki);
den_G_a=((La*Jm*Jl)*s^5+(Jl*Ra*Jm+Jl*Bm*La)*s^4+(Ki*Kb*Jl+Ra*Bm*Jl)*s
^3);
figure(1);
rlocus(G_a)
%b)
K=1000;
num_G_b=((n^2*La*Jl+La*Jm)*s^3+(n^2*Ra*Jl+Ra*Jm+Bm*La)*s^2+Ra*Bm*s+Ki
*Kb*s+n*Ks*K*Ki);
den_G_b=((La*Jm*Jl)*s^5+(Jl*Ra*Jm+Jl*Bm*La)*s^4+(Ki*Kb*Jl+Ra*Bm*Jl)*s
^3);
figure(2);
rlocus(G_b)

Root Locus
1000
800
600
400
200
Imaginary Axis
-200
-400
-600
-800
-1000
-3500 -3000 -2500 -2000 -1500 -1000 -500 0 500
Real Axis
7130
Root Locus
1000
800
600
400
200
Imaginary Axis
-200
-400
-600
-800
-1000
-4000 -3500 -3000 -2500 -2000 -1500 -1000 -500 0 500
Real Axis
7-42
(a) CharacteristicEquation: s + 5000 s + 572,400 s + 900,000 + J L 10 s + 50,000 s
3 2
( 3 2
)= 0
3 2 2
P ( s ) = s + 5000 s + 572, 400 s + 900,000 = ( s + 1.5945)( s + 115.6 )( s + 4882.8 ) Q ( s ) = 10 s ( s + 5000 )
Sincethepoleat5000isveryclosetothezeroat4882.8, P ( s ) and Q( s ) canbeapproximatedas:
2
P ( s ) ( s + 1.5945)( s + 115.6 ) Q ( s ) 10.24 s
2
BreakawaypointEquation: 1200 s + 3775 s = 0 BreakawayPoints:(RL):0,3.146
7131
(b) MATLAB code:

s = tf('s')
K=1;
Jm=0.001;
La=0.001;
n=0.1;
Ra=5;
Ki=9;
Bm=0;
Kb=0.0636;
Ks=1;
num_G_a = (n^2*La*s^3+n^2*Ra*s^2);
den_G_a = (La*Jm*s^3+(Ra*Jm+Bm*La)*s^2+(Ra*Bm+Ki*Kb)*s+n*Ki*Ks*K);
figure(1);
rlocus(G_a)
Root locus diagram:
Root Locus
300
200
100
Imaginary Axis
-100
-200
-300
-6000 -5000 -4000 -3000 -2000 -1000 0 1000
Real Axis
7132
7-43)(a) = 12: P ( s ) = s 2 ( s + 12 ) Q ( s ) = s + 1
o o o o
Asymptotes: K>0: 90 , 270 K<0: 0 , 180
0 + 0 12 ( 1)
1 = = 5.5
31
3 2
BreakawaypointEquation: 2 s + 15 s + 24 s = 0 BreakawayPoints:0,2.314,5.186
2
743(b) = 4: P( s) = s ( s + 4) Q ( s ) = s + 1
o o o o
Asymptotes:K>0: 90 , 270 K<0: 0 , 180
0 + 0 4 ( 10
1 = = 15
.
31
3 2
BreakawaypointEquation: 2 s + 7 s + 8 s = 0 BreakawayPoints:K>00.NoneforK<0.
7133
2
2 +3 ( + 3) 16
(c)BreakawaypointEquation: 2 s + ( + 3) s + 2 s = 0 Solutions: s = , s = 0
4 4
Foronenonzerobreakawaypoint,thequantityunderthesquarerootsignmustequalzero.
2
Thus, 10 + 9 = 0, = 1 or = 9. The answer is = 9. The = 1solutionrepresentspolezero
cancellationintheequivalent G ( s ). When = 9, thenonzerobreakawaypointisat s = 3. 1 = 4.
7134
7-44)
For part (c), after finding the expression for:
dk 3 ( 1)( 9)
= ,
ds 4
there is one acceptable value of alpha that makes the square root zero ( = 9 ). Zero square root means
one answer to the breakaway point instead of 2 answers as a result of sign. = 1 is not acceptable
dk 0
since it results in s = 1@ = 0 and then k = .
ds 0
MATLAB code:
s = tf('s')
%(a)
alpha=12
num_GH= s+1;
den_GH=s^3+alpha*s^2;
GH=num_GH/den_GH;
figure(1);
rlocus(GH)
%(b)
alpha=4
num_GH= s+1;
GH=num_GH/den_GH;
figure(2);
rlocus(GH)
%(c)
alpha=9
num_GH= s+1;
GH=num_GH/den_GH;
figure(3);
rlocus(GH)
7135
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5
Real Axis
7136
Root locus diagram, part (c): (alpha=9 resulting in 1 breakaway point)
Root Locus
8
2
Imaginary Axis
-2
-4
-6
-8
-10 -8 -6 -4 -2 0 2
Real Axis
7-45)(a) P ( s ) = s 2 ( s + 3) Q( s ) = s +
3
BreakawaypointEquation: 2 s + 3(1 + ) s + 6 = 0
Therootsofthebreakawaypointequationare:
2
3(1 + ) 9(1 + ) 48
s=
4 4
2
Fornobreakawaypointotherthanat s = 0 ,set 9(1 + ) 48 < 0 or 0.333 < < 3
RootLocusDiagramwithNoBreakawayPointotherthanats=0.
7137
745(b)Onebreakawaypointotherthanats=0: = 0.333, Breakaway point at s = 1.
7138
7139
745(d)Twobreakawaypoints: > 3 :
7-46) First we can rearrange the system as:
H 1 (s) e st
where
Now designing a controller is similar to the designing a controller for any unity feedback system.
7140
7-47) Lettheangleofthevectordrawnfromthezeroat s = j12 toapoint s1 ontherootlocussnearthezero
be . Let
1 = angle of the vector drawn from the pole at j10 to s1 .

2 = angle of the vector drawn from the pole at 0 to s1 .

3 = angle of the vector drawn from the pole at j10 to s1 .
4 = angle of the vector drawn from the zero at j12 to s1 .
Thentheangleconditionsontherootlociare:
= 1 2 3 + 4 = odd multiples of 180

o

o
1 = 2 = 3 = 4 = 90 =0
o
Thus,

Therootlocishownin(b)arethecorrectones.
Answers to True and False Review Questions:
6. (F) 7. (T) 8. (T) 9. (F) 10. (F) 11. (T) 12. (T) 13. (T) 14. (T)
7141
Chapter 8
6.54
81(a)K=5 n = 5 = 2.24 rad / sec = = 1.46 Mr = 1 r = 0 rad / sec
4.48
6.54 1
(b)K=21.39 n = 2139
. = 4.62 rad / sec = = 0.707 M r = =1
9.24 2
2 1
2
r = n 1 = 3.27 rad / sec
6.54
(c)K=100 n = 10 rad / sec = = 0.327 M r = 1618
. r = 9.45 rad / sec
20
8-2 Bode diagram (a) k=5: data points from top to bottom indicate
MATLAB code: bandwidth BW, resonance peak Mr, and resonant frequency r.
% Question 8-2, System: CL_a
clear all; Frequency (rad/sec): 0.855
Bode Diagram Magnitude (dB): -3
close all; 0
s = tf('s') -10
System: CL_a
Magnitude (dB)
-20 Frequency (rad/sec): 2.25

%a) Magnitude (dB): -9.4
num_G_a= 5; -30
den_G_a=s*(s+6.54); -40
-50
CL_a=G_a/(1+G_a)
0
BW = bandwidth(CL_a)
bode(CL_a)
-45
Phase (deg)
%b)
figure(2); -90
System: CL_a
num_G_b=21.38; Frequency (rad/sec): 2.25
den_G_b=s*(s+6.54); -135 Phase (deg): -90.2
-1 0
10 10
G_b=num_G_b/den_G_b; Frequency (rad/sec)
CL_b=G_b/(1+G_b)
BW = bandwidth(CL_b) Bode diagram (b) k=21.38: data points from top to bottom
bode(CL_b)
indicate bandwidth BW, resonance peak Mr, and resonant frequency
%c) r.
figure(3);
num_G_c=100;
den_G_c=s*(s+6.54);
81
CL_c=G_c/(1+G_c) System: CL_b

BW = bandwidth(CL_c) Magnitude (dB): -3
Bode Diagram
0
bode(CL_c)
System: CL_b
Magnitude (dB)
Magnitude (dB): -3.03
-40
-60
-80
-100
0
-45
Phase (deg)
-90
System: CL_b
Phase (deg): -89.9
-180
0 1 2
10 10 10
Frequency (rad/sec)
Bode diagram (c) k=100: data points from top to bottom indicate
resonance peak Mr, bandwidth BW, and resonant frequency r.
Bode Diagram
20
0
System: CL_c
Magnitude (dB)
Frequency (rad/sec): 9.99 System: CL_c

Magnitude (dB): 3.62
-40
-60
-80
0
-45
Phase (deg)
-90
System: CL_c
-135 Frequency (rad/sec): 10
Phase (deg): -90
-180
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
82
8-3) If is the input, then
where
and
Therefore:
As a result:
8-4(a) Mr = 2.944 ( 9.38 dB) r = 3 rad / sec BW = 4.495 rad / sec
(b) M r = 15.34 ( 23.71 dB) r = 4 rad / sec BW = 6.223 rad / sec
(c) M r = 4.17 (12.4 dB) r = 6.25 rad / sec BW = 9.18 rad / sec
(d) M r = 1 ( 0 dB) r = 0 rad / sec BW = 0.46 rad / sec
(e) M r = 157
. ( 3.918 dB) r = 0.82 rad / sec BW = 1.12 rad / sec
83
(f) Mr = ( unstable) r = 15
. rad / sec BW = 2.44 rad / sec
(g) M r = 3.09 ( 9.8 dB) r = 1.25 rad / sec BW = 2.07 rad / sec
(h) M r = 4.12 (12.3 dB) r = 3.5 rad / sec BW = 5.16 rad / sec
85)
Maximum overshoot = 0.1 Thus, = 0.59
2
1 1 0.416 + 2.917
Mr = = 105
. tr = = 0.1 sec
2 1
2 n
Thus,minimum n = 17.7 rad / sec Maximum M r = 105

.
((1 2 )+ )
1/ 2
MinimumBW= n 4 4 + 2 = 20.56 rad/sec

2 4 2
86)

2
1
Maximum overshoot = 0.2 Thus, 0.2 = e = 0.456
2
1 1 0.416 + 2.917
Mr = = 1.232 t r = = 0.2 Thus,minimum n = 14.168 rad/sec
2 1
2 n
(( )+ )
1/ 2
Maximum M r = 1.232 MinimumBW= 1 2 4 4 + 2 = 18.7 rad/sec

2 4 2

2
1
8-7) Maximumovershoot=0.3 Thus, 0.3 = e = 0.358
2
1 1 0.416 + 2.917
Mr = = 1.496 t r = = 0.2 Thus,minimum n = 6.1246 rad/sec
2 1
2 n
84
(( )+ )
1/ 2
Maximum M r = 1.496 MinimumBW= 1 2 4 4 + 2 = 1.4106 rad/sec

2 4 2
8-8) (a)
At the gain crossover:
Therefore:
at = 1.5
(b)
MATLAB code:
%solving for k:
syms kc
omega=1.5
sol=eval(solve('0.25*kc^2=0.7079^2*((-0.25*omega^3+omega)^2+(-
0.375*omega^2+0.5*kc)^2)',kc))
%ploting bode with K=1.0370
s = tf('s')
K=1.0370;
num_G_a= 0.5*K;
den_G_a=s*(0.25*s^2+0.375*s+1);
CL_a = G_a/(1+G_a)
bode(CL_a);
Bode diagram: data point shows -3dB point at 1.5 rad/sec frequency which is the closed loop
bandwidth
85
Bode Diagram
0
System: CL_a
-20
Magnitude (dB) -40
-60
-80
-100
-120
0
-90
Phase (deg)
-180
-270
-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/sec)
8-9) =
= 90 = 63o
Therefore:
As a result:
86
Therefore:
To change the crossover frequency requires adding gain as:
(b) MATLAB code:

s = tf('s')
%(b)
K = 0.95*2;
num_G_a = 0.5*K;
den_G_a = s*(0.25*s^2+0.375*s+1);
CL_a = G_a/(1+G_a)
bode(CL_a);
figure(2);
sisotool
Peak mag = 2.22 can be converted to dB units by: 20*Log(2.22)= 6.9271 dB
87
By using sisotool and importing the loop transfer function, the overall gain (0.5K) was changed until the
magnitude of the resonance in Bode was about 6.9 dB. At 0.5K=~0.95 or K=1.9, this resonance peak
was achieved as can be seen in the BODE diagram of the following figure:
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
6 20
4 0
2 -20
0 -40
-2 -60
-4 -80
G.M.: 4.48 dB
-6 -100 Freq: 2 rad/sec
-8 -6 -4 -2 0 2 4
Stable loop
Bode Editor for Closed Loop 1 (CL1) -120
-90
6
4 -135
2
-180
0
-225
-180
P.M.: 59.1 deg
Freq: 1.11 rad/sec
-360 -270
-2 -1 0 1 2 -1 0 1 2
10 10 10 10 10 10 10 10 10
Frequency (rad/sec) Frequency (rad/sec)
8-10)

1 1
2
M r = 1.4 = Thus, = 0.387Maximumovershoot= e = 0.2675 (26.75%)
2
2 1
2 3
r = 3 rad / sec = n 1 2 = 0.8367 n rad/sec n = = 3.586rad/sec
0.8367

t max = = = 0.95 sec At = 0, M = 0.9.
2 2
n 1 3.586 1 ( 0.387 )
88
Thisindicatesthatthesteadystatevalueoftheunitstepresponseis0.9.
UnitstepResponse:
8-11) a) The closed loop transfer function is:
as ,which means = 0.387
According to the transfer function: n = 0.1 n = 0.129 rad/s

As n2 = 0.1K ; then, K = 10 n2 = 0.1669
b)
As K = 0.1664, then
which means
Accordingly PM = 42o
As , then
89
8-12)
T BW(rad/sec) Mr
________________________________________________________________
0 1.14 1.54
0.5 1.17 1.09
1.0 1.26 1.00
2.0 1.63 1.09
3.0 1.96 1.29
4.0 2.26 1.46
5.0 2.52 1.63
_________________________________________________________________
813)
T BW(rad/sec) Mr
_________________________________________________________________
0 1.14 1.54
0.5 1.00 2.32
1.0 0.90 2.65
2.0 0.74 2.91
3.0 0.63 3.18
4.0 0.55 3.37
5.0 0.50 3.62
_________________________________________________________________
8-14) The Routh array is:
810
S3 0.25 1
S2 0.375 0.5K
S1 1-1/3 0
S0 0.5K
Therefore:
As , if GH is rearranged as:
then
which gives
where
therefore,
(c)
MATLAB code: Bode diagram:
s = tf('s')
%c)
K = 1.03697;
num_G_a = 0.5*K;
den_G_a =
s*(0.25*s^2+0.375*s+1
);
%create closed-loop
system
G_a =
num_G_a/den_G_a;
811
CL_a = G_a/(1+G_a)
bode(CL_a); Bode Diagram
0
System: CL_a
-20
Magnitude (dB)
Notes: -40
-60
1 - BW is verified by finding - -80

3dB point at Freq = 1.5 rad/sec -100
in the Bode graph at calculated -120
k. 0
2- By comparison to diagram of Phase (deg)

-90
typical 2nd order poles with

different damping ratios, -180
damping ratio is approximated

as: -270
-2 -1 0 1 2
= ~ .707 10 10 10 10 10
Frequency (rad/sec)
812
815(a)
20
L( s ) = P = 1, P = 0
s(1 + 0.1s )(1 + 0.5s )
When = 0: L( j ) = 90o L( j ) = When = : L( j ) = 270

o
L( j ) = 0
L ( j ) =
20
=
20 0.6 j 1 0.05
2
( 2
) Setting Im L( j ) = 0
2
(
0.6 + j 1 0.05
2
) 0.36 +
4 2
(1 0.05 )
2 2
2
1 0.05 = 0 Thus, = 4.47 rad / sec L( j 4.47 ) = 1667
.
360
o
11 = 270 = ( Z 0.5 P P ) 180 = ( Z 0.5 ) 180 Thus, Z = = 2

o o o
180
o
Theclosedloopsystemisunstable.Thecharacteristicequationhastworootsintherighthalf
splane.
MATLAB code:
s = tf('s')
%a)
figure(1);
num_G_a= 20;
den_G_a=s*(0.1*s+1)*(0.5*s+1);
nyquist(G_a)
NyquistPlotof L( j ):
813
(b)
10
L( s ) = Basedontheanalysisconductedinpart(a),theintersectofthenegative
s(1 + 0.1s )(1 + 0.5s )
realaxisbythe L( j ) plotisat0.8333,andthecorresponding is4.47rad/sec.

o
o
11 = 90 = Z 0.5 P P 180 = 180 Z 90
o
Thus, Z = 0. Theclosedloopsystemisstable.
MATLAB code:
s = tf('s')
%b)
figure(1);
num_G_a= 10;
den_G_a=s*(0.1*s+1)*(0.5*s+1);
nyquist(G_a)
814
(c)
100(1 + s )
L( s ) = P = 1, P = 0.
s(1 + 0.1s )(1 + 0.2 s )(1 + 0.5s )
When = 0: L( j 0) = 90o L( j 0) = When = : L( j ) = 270 o L( j ) = 0
When = : L( j ) = 270 L( j ) = 0 When = : L( j ) = 270 L( j ) = 0

o o
L ( j ) =
100(1 + j )
=
(
100(1 + j ) 0.01 0.8
4 2
) j (1 0.17 )
2
( 0.01 4
0.8
2
) + j (1 0.17 ) 2
( 0.01 4
0.8
2
) + (1 0.17 )
2 2 2 2
4 2 2 4 2
Setting Im L( j ) = 0 0.01 0.8 1 + 0.17 = 0 63 100 = 0
2
Thus, = 64.55 = 8.03 rad/sec
100 ( 0.01 4 0.8 2 ) + 2 (1 0.17 2 )

L( j8.03) = = 10
( 0.01 2 0.8 2 )2 + 2 (1 0.17 2 )2
=8.03
815
11 = 270 = ( Z 0.5 P P ) 180 = ( Z 0.5 ) 180 Thus, Z = 2 Theclosedloopsystemis

o o o
unstable.
Thecharacteristicequationhastworootsintherighthalfsplane.
MATLAB code:
s = tf('s')
%c)
figure(1);
num_G_a= 100*(s+1);
den_G_a=s*(0.1*s+1)*(0.2*s+1)*(0.5*s+1);
nyquist(G_a)
816
(d)
10
L( s ) = 2
P = 2 P = 0
s (1 + 0.2 s )(1 + 0.5s )
When = 0: L( j ) = 180o L( j ) = When = : L( j ) = 360

o
L( j ) = 0
L ( j ) =
10
=
(
10 0.1 + j 0.7
4 2 3
)
( 0.1 4

2
) j 0.7 3
( 0.1 4

2
)
2
+ 0.49
6
Setting Im L( j ) = 0, = . TheNyquistplotof L( j ) doesnotintersecttherealaxisexceptatthe
originwhere = .
11 = ( Z 0.5 P P ) 180 = ( Z 1) 180 Thus, Z = 2.

o o

Theclosedloopsystemisunstable.Thecharacteristicequationhastworootsintherighthalfsplane.
MATLAB code:
s = tf('s')
%d)
figure(1);
num_G_a= 10;
den_G_a=s^2* (0.2*s+1)*(0.5*s+1);
nyquist(G_a)
817
815(e)
3( s + 2)
L( s ) = P = 1 P = 2
(
s s + 3s + 1
3
)
When = 0: L( j 0) = 90o L( j 0 ) = When = : L( j ) = 270

o
L( j ) = 0
L ( j ) =
3( j + 2)
=
3( j + 2) 3( 4 2
) j Setting Im L( j ) = 0,
( 4
3
2
) + j (4 4
3
2
)
2
+
2
4 2 2
3 2 = 0 or = 3.56 = 189
. rad/sec. . ) = 3
L( j189
11 = ( Z 0.5 P P ) 180 = ( Z 2.5 ) 180 = 90 Thus, Z = 2

o o o

MATLAB code:
818
s = tf('s')
%e)
figure(1);
num_G_a= 3*(s+2);
den_G_a=s*(s^3+3*s+1);
nyquist(G_a)
815(f)
0.1
L( s ) = P = 1 P = 0
(
s ( s + 1) s + s + 1
2
)
When = 0: L( j 0) = 90 When = : L( j ) = 360

o o
L( j 0 ) = L( j ) = 0
L ( j ) =
0.1
=
(
0.1 2
4 2
) j (1 2 ) 2
Settiing Im L( j ) = 0
( 4
2
2
) + j (1 2 ) 2
( 4
2
2
) + (1 2 )
2 2 2 2
2
= or = 0.5 = 0.707 rad / sec L( j 0.707 ) = 0.1333
11 = ( Z 0.5 P P ) 180 = ( Z 0.5 ) 180 = 90 Thus, Z = 0 Theclosedloopsystemisstable.

o o o
819
MATLAB code:
s = tf('s')
%f)
figure(1);
num_G_a= 0.1;
den_G_a=s*(s+1)*(s^2+s+1);
nyquist(G_a)
815(g)
100
L( s ) = P = 3 P = 0
(
s ( s + 1) s + 2
2
)
When = 0: L( j 0) = 90 When = : L( j ) = 360
o o
L( j 0 ) = L( j ) = 0
Thephaseof L( j ) isdiscontinuousat = 1.414 rad/sec.
o
(
11 = 35.27 + 270 215.27
o o
) = 90
o
11 = ( Z 1.5 ) 180 = 90 Thus, P11 =
o o 360
o
= 2
180
820
MATLAB code:
s = tf('s')
%g)
figure(1);
num_G_a= 100;
den_G_a=s*(s+1)*(s^2+2);
nyquist(G_a)
815(h)
10( s + 10)
L( s ) = P = 1 P = 0
s( s + 1)( s + 100)
When = 0: L( j 0) = 90o L( j 0 ) = When = : L( j ) = 180

o
L( j ) = 0
L ( j ) =
10( j + 10)
=
(
10( j + 10) 101 j 100
2 2
)
101 + j (100 ) (100 )
2 2 2
10201 +
4 2 2
821
Setting Im L( j ) = 0, = 0 istheonlysolution.Thus,theNyquistplotof L( j ) doesnotintersect
therealaxis,exceptattheorigin.
11 = ( Z 0.5 P P ) 180 = ( Z 0.5 ) 180 = 90 Thus, Z = 0.

o o o
Theclosedloopsystemisstable.
MATLAB code:
s = tf('s')
%h)
figure(1);
num_G_a= 10*(s+10);
den_G_a=s*(s+1)*(s+100);
nyquist(G_a)

822
8-16
MATLAB code:
s = tf('s')
%a)
figure(1);
num_G_a= 1;
den_G_a=s*(s+2)*(s+10);
nyquist(G_a)
%b)
figure(2);
num_G_b= 1*(s+1);
den_G_b=s*(s+2)*(s+5)*(s+15);
nyquist(G_b)
%c)
figure(3);
num_G_c= 1;
den_G_c=s^2*(s+2)*(s+10);
nyquist(G_c)
%d)
figure(4);
num_G_d= 1;
den_G_d=(s+2)^2*(s+5);
823
nyquist(G_d)
%e)
figure(5);
num_G_e= 1*(s+5)*(s+1);
den_G_e=(s+50)*(s+2)^3;
nyquist(G_e)
Nyquist graph, part(a):
Nyquist Diagram
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
Nyquist graph, part(b):
824
Nyquist Diagram
4
1
Imaginary Axis
-1
-2
-3
-4
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Real Axis
Nyquist graph, part(c):
Nyquist Diagram
0.5
0.4
0.3
0.2
0.1
Imaginary Axis
-0.1
-0.2
-0.3
-0.4
-0.5
-10 -8 -6 -4 -2 0 2
Real Axis
Nyquist graph, part(d):
825
Nyquist Diagram
0.04
0.03
0.02
0.01
Imaginary Axis
-0.01
-0.02
-0.03
-0.04
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Real Axis
Nyquist graph, part (e):
Nyquist Diagram
0.01
0.008
0.006
0.004
0.002
Imaginary Axis
-0.002
-0.004
-0.006
-0.008
-0.01
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Real Axis
826
8-17(a)
K
G ( s ) = 2
P = 0 P = 0
( s + 5)
o K
G ( 0) = 180 ( K < 0) G ( j 0) =
o
G ( j 0 ) = 0 ( K > 0 )
25
G ( j ) = 0
o o
G ( j ) = 180 (K>0) G ( j ) = 0 (K<0)
Forstability,Z=0.
11 = ( 0.5 P + P ) 180 = 0
o o
o
0 < K < 11 = 0 Stable
o
K < 25 11 = 180 Unstable
o
25 < K < 0 11 = 0 Stable

Thesystemisstablefor25<K< .
817(b)
K
G ( s ) = 3
P = 0 P = 0
( s + 5)
o K
G ( 0) = 180 ( K < 0) G ( j 0) =
o
G ( j 0 ) = 0 ( K > 0 )
125
G ( j ) = 270 (K<0) G ( j ) = 0
o o
G ( j ) = 270 (K>0)
827
Forstability,Z=0.
11 = ( 0.5 P + P ) 180 = 0
o o
o
0<K<1000 11 = 0 Stable
o
K > 1000 11 = 360 Unstable
o
K < 125 11 = 180 Unstable
o
125 < K < 0 11 = 0 Stable
Thesystemisstablefor125<K<0.
817(c)
K
G ( s ) = 4
P = P = 0
( s + 5)
o K
G ( 0 ) = 180 ( K < 0) G ( j 0) =
o
G ( j 0 ) = 0 ( K > 0 )
625
G ( j ) = 180 (K<0) G ( j ) = 0
o o
G ( j ) = 0 (K>0)
828
Forstability,Z=0.
11 = ( 0.5 P + P ) 180 = 0
o o
o
0 < K < 2500 11 = 0 Stable
o
K > 2500 11 = 360 Unstable
o
K < 625 11 = 180 Unstable

o
625 < K < 0 11 = 0 Stable
Thesystemisstablefor625<K<2500.
8-18) The characteristic equation:
or
if , the cross real axis at s = 0.1. For stability
if K < 0, the Nyquist cross the real axis at .So, for stability,
therefore, the rage of stability for the system is
MATLAB code:
s = tf('s')
K=1
G= K/(s^2+2*s+2);
H=1/(s+1);
GH=G*H;
sisotool
829
K=10.15
G2= K/(s^2+2*s+2);
H2=1/(s+1);
GH2=G2*H2;
nyquist(GH2)
xlim[(-1.5,.5)]
ylim[(-1,1)]
After generating the feed-forward (G) and feedback (H) transfer functions in the MATLAB code, these
transfer functions are imported to sisotool. Nyquist diagram is added to the results of sisotool. The
overall gain of the transfer function is changed until Nyquist diagram passes through -1+0j point. Higher
values of K resulted in unstable Nyquist diagram. Therefore K<10.15 determines the range of stability
for the closed loop system.
Nyquist at margin of stability:
Nyquist Diagram
4
1
Imaginary Axis
-1
-2
-3
-4
-2 -1 0 1 2 3 4 5 6
Real Axis
8-19)
( )
s s + 2s + s + 1 + K s + s + 1 = 0
3 2
( 2
)
830
Leq ( s ) =
(
K s + s +1
2
) P = 1 P = 0 Leq ( j 0 ) = 90
o
Leq ( j ) = 0180
o

(
s s + 2s + s + 1
3 2
)
Leq ( j ) =
(
K 1 ) + j = K (
2 6
+
4
) j ( 2 + 1)
4 2
( 4

2
) + j (1 2 ) 2
( 4

2
) + (1 2 )
2 2 2 2
Setting Im Leq ( j ) = 0
4 2
2 + 1 = 0
Thus, = 1rad/secaretherealsolutions.
Leq ( j1) = K
Forstability,
o o

11 = ( 0.5P + P )180 = 90
WhenK=1thesystemismarginallystable.
o
K > 0 11 = 90 Stable
o
K < 0 11 = +90 Unstable
831
RouthTabulation
4
s 1 K +1 K
3
s 2 K +1
2 K +1
s K K > 1
2
2 2
1 K 2K + 1 ( K 1)
s =
K +1 K +1
0
s K K>0
1 2
WhenK=1thecoefficientsofthe s rowareallzero.Theauxiliaryequationis s + 1 = 0 Thesolutions
are = 1rad/sec.ThustheNyquistplotof Leq ( j ) intersectsthe1pointwhenK=1,when = 1
rad/sec.Thesystemisstablefor0<K< ,exceptatK=1.
8-20) Solution is similar to the previous problem. Lets use Matlab as an alternative approach
MATLAB code:
s = tf('s')
figure(1);
K=8.09
num_GH= K;
den_GH=(s^3+3*s^2+3*s+1);
GH=num_GH/den_GH;
nyquist(GH)
xlim([-1.5,.5])
ylim([-1,1])
sisotool;
832
After generating the loop transfer function and analyzing Nyquist in MATLAB sisotool, it was found
that for values of K higher than ~8.09, the closed loop system is unstable. Following is the Nyquist
diagram at margin of stability.
Part(a), Nyquist at margin of stability:
Nyquist Diagram
1
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1
-1.5 -1 -0.5 0 0.5
Real Axis
Part(b), Verification by Routh-Hurwitz criterion:

Using Routh criterion, the coefficient table is as follows:
S3 1 3
S2 3 K+1
S1 (8-K)/3 0
S0 K+1 0
The system is stable if the content of the 1st column is positive:

(8-K)/3>0 K<8
833
K+1>0 K>-1
which is consistent with the results of the Nyquist diagrams.
8-21)
Parabolicerrorconstant K a = lim s G ( s ) = lim10 ( K P + K D s ) = 10 K P = 100 Thus K P = 10

2
s 0 s 0
2
CharacteristicEquation: s + 10 K D s + 100 = 0
10 KD s
Geq ( s ) = 2
P = 2 P = 0
s + 100
Forstability,
o o
11 = ( 0.5 P + P )180 = 180
Thesystemisstablefor 0 < K D < .
834
822(a)Thecharacteristicequationis1 + G ( s ) G ( s ) 2 G ( s ) 2 = 1 2 G ( s ) 2 = 0
2
2 2 K
Geq ( s ) = 2 G ( s ) = 2 2
P = 0 P = 0
( s + 4 ) ( s + 5)
Geq ( j ) =
2 K
2
=
2
(
2 K 400 120 +
2 2
) j ( 360 18 )
2
( 400 120 2
+
4
) + j ( 360 18 ) 2
( 400 120 2
+
2
) + ( 360 18 )
2 2 2
2
K
Geq ( j 0) = 180
o
Geq ( j ) = 0180
o
Setting Im Geq ( j ) = 0
200
2
K
= 0 and = 4.47 rad / sec Geq ( j 4.47 ) =
800
Forstability,
11 = ( 0.5 P + P ) 180 = 0
o o
2
Thesystemisstablefor K < 200
or K < 200

835
822(b)
4 3 2 2
CharacteristicEquation: s + 18 s + 121s + 360 s + 400 2 K = 0
RouthTabulation
4 2
s 1 121 400 2 K
3
s 18 360
2 2
s 101 400 2 K
2
1 29160 - 36 K 2
s 29160 + 36 K > 0
101
s 0 400 2 K 2 K 2 < 200
Thusforstability, K < 200
823(a)
NyquistPlot
83.33N
G ( s ) =
s( s + 2)( s + 11767
. )
Forstability,N<3.89
ThusN<3sinceNmustbeaninteger.
836
(b)
2500
G ( s) =
s( 0.06 s + 0.706)( As + 100)
CharacteristicEquation: 0.06 As 3 + ( 6 + 0.706 A) s 2 + 70.6 s + 2500 = 0

2
As ( 0.06 s + 0.706 )
Geq ( s ) = 2
SinceGeq ( s ) hasmorezerosthanpoles,weshouldsketchtheNyquist
6 s + 70.6 s + 2500
plotof1/ Geq ( s ) forstabilitystudy.

1
=
( 2500 6 ) + j 70.6 = ( 2500 6 ) + j 70.6 ( 0.706 + j 0.06 )
2 2 2 3
Geq ( j ) A ( 0.706 j 0.06 ) A ( 0.498 + 0.0036 )

2 3 4 6
1
= 0
o o
1 / Geq ( j 0 ) = 180 1 / Geq ( j ) = 0 90 Setting Im
Geq ( j )
2 1 4.23
100.156 0.36 = 0 = 16.68 rad/sec =
Geq ( j16.68) A
Forstability,
11 = ( 0.5 P + P ) 180 = 180

o o
o
For A > 4.23 11 = 180 Unstable
o
For 0 < A < 4.23 11 = 180 Stable
Thesystemisstablefor0<A<4.23.
(c)
837
2500
G ( s ) =
s( 0.06 s + 0.706 )( 50 s + Ko )
CharacteristicEquation: s( 0.06 s + 0.706)(50 s + Ko ) + 2500 =0
Ko s( 0.06 s + 0.706)
Geq ( s ) = P = 0 P = 0 Geq ( j 0 ) = 0 90
o o
3 2
Geq ( j ) = 0 90
3s + 35.3s + 2500
Geq ( j ) =
K o ( 0.06 + 0.706 j )
3
=
( )(
K o 0.06 + 0.706 j 2500 35.5
2 2
) + j3 3

( 2500 35.3 ) j3
2 3
( 2500 35.3 ) 2 2
+ 9
6
4 2 2
Setting Im Geq ( j ) = 0 + 138.45 9805.55 = 0 = 516
. = 7.18 rad/sec
Geq ( j 7.18 ) = 0.004 K o
Forstability, 11 = ( 0.5 P + P ) 180 = 0

o o
Forstability, 0 < Ko < 217.4
838
824(a) Kt = 0:
Y (s) 10000 K 10000 K

G(s) = = =
E ( s) (
s s + 10 s + 10000 K t
2
) s ( s + 10)
2
The(1,j0)pointisenclosedforall
valuesofK.Thesystemisunstable
forallvaluesofK.
(b) Kt = 0.01:
G(s) =
10000 K
G ( j ) =
10000 K 10 j 100
2
( 2
)
(
s s + 10 s + 100
2
) 100 +
4 2
(100 ) 2 2
Setting Im G ( j ) = 0 2 = 100
= 10 rad / sec G ( j10 ) = 10 K
Thesystemisstablefor0<K<0.1
839
(c) Kt = 0.1:
G( s) =
10000 K
G ( j ) =
10000 K 10 j 1000
2
( 2
)
(
s s + 10 s + 1000
2
) 100 +
4 2
(1000 ) 2 2
Setting Im G ( j ) = 0 2 = 100 = 31.6 rad/sec . ) = K

G ( j 316
Forstability,0<K<1
825)ThecharacteristicequationforK=10is:
3 2
s + 10 s + 10,000 Kt s + 100,000 = 0
10,000 Kt s
Geq ( s ) = 3 2
P = 0 P = 2
s + 10 s + 100,000
Geq ( j ) =
10, 000 K t j
=
(
10, 000 K t + j 10, 000 10
4 2
) Setting Im G ( j ) = 0
100, 000 10 j (10, 000 10 )
2 3 2 eq
+
2 6
840
2
= 0, = 10,000
= 100 rad / sec Geq ( j100) = Kt
Forstability,
11 = ( 0.5 P + P ) 180 = 360

o o

Thesystemisstablefor Kt > 0.
8-26)
a)
b)
c)
MATLAB code:
s = tf('s')
figure(1);
J=1;
B=1;
K=1;
Kf=0
G1= K/(J*s+B);
CL1=G1/(1+G1*Kf);
H2 = 1;
G1G2 = CL1/s;
L_TF=G1G2*H2;
nyquist(L_TF)
sisotool
841
Part (a), Kf=0: by plotting the Nyquist diagram in sisotool and varying the gain, it was observed that
all values of gain (K) will result in a stable system. Location of poles in root locus diagram of the second
figure will also verify that.
Nyquist Diagram
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
842
1 40
20
0.5
0
0
-20
-0.5
-40
G.M.: Inf
-1 -60
Freq: Inf
-1 -0.8 -0.6 -0.4 -0.2 0
Stable loop
50
-90
-50
-100 -135
0
-90
P.M.: 54.9 deg
Freq: 0.704 rad/sec
-180 -180
-2 -1 0 1 2 -2 -1 0 1 2
10 10 10 10 10 10 10 10 10 10
Part (b), Kf=0.1: The result and approach is similar to part (a), a sample of Nyquist diagram is
presented for his case as follows:
Nyquist Diagram
15
10
5
Imaginary Axis
-5
-10
-15
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
843
Part (c), Kf=0.2: The result and approach is similar to part (a), a sample of Nyquist diagram is
presented for his case as follows:
Nyquist Diagram
15
10
5
Imaginary Axis
-5
-10
-15
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
8-27)
MATLAB code:
s = tf('s')
figure(1);
J=1;
B=1;
K=10;
Kf=0.2
G1= K/(J*s+B);
CL1=G1/(1+G1*Kf);
H2 = 1;
G1G2 = CL1/s;
L_TF=G1G2*H2;
nyquist(L_TF)
After assigning K=10, different values of Kf has been used in the range of 0.01<K<104. The Nyquist
diagrams shows the stability of the closed loop system for all 0<K<. A sample of Nyquist diagram is
plotted as follows:
844
Nyquist Diagram
25
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-25
-1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0
Real Axis
8-28) a) K > 2 system is stable
b) 0 < K < 1 and -2 < K < 0 -2 < K < 1 system is stable
MATLAB code:
s = tf('s')
%a)
figure(1);
K=1
num_GH_a= K*(s+1);
den_GH_a=(s-1)^2;
nyquist(GH_a)
%b)
figure(2);
K=1
num_GH_b= K*(s-1);
den_GH_b=(s+1)^2;
nyquist(GH_b)
845
sisotool
Part(a): Using MATLAB sisotool, the transfer function gain can be iteratively changed in order to
obtain different phase margins. By changing the gain so that PM=0 (margin of stability), K>~2 resulted
in stable Nyquist diagram for part(a). Following two figures illustrate the sisotool and Nyquist results at
margin of stability for part (a).
2 10
1 5
0
0
-5
-1
-10 G.M.: -0.00581 dB

-2 Freq: 1.73 rad/sec
-4 -3 -2 -1 0 1
Stable loop
100
-90
P.M.: 0.0688 deg
50
0 -180
-50 -225
180
-270
0
-315
-180 -360
-2 -1 0 1 2 0
10 10 10 10 10 10
846
Nyquist Diagram
2
1.5
Imaginary Axis 0.5
-0.5
-1
-1.5
-2
-1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Real Axis
Part(b): Similar methodology applied as in part (a). K<1 results in closed loop stability. Following are
sisotool and Nyquist results at margin of stability (K=1):
1 0
0.5
-5
0
-10
-0.5
-15
G.M.: -0.00236 dB
-1 Freq: 0 rad/sec
-5 -4 -3 -2 -1 0 1 2
Unstable loop
80
180
60
40
135
20
0 90
-20
-40 45
90
45 0
0
-45 P.M.: -4.01 deg
-45
Freq: 0.0233 rad/sec
-90 -90
-6 -4 -2 0 2 -2 -1 0 1
10 10 10 10 10 10 10 10 10
847
Nyquist Diagram
1
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1
-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6
Real Axis
8-29)(a)
Td s 100
Let G ( s ) = G1 ( s )e Then G1 ( s ) =
(
s s + 10 s + 100
2
)
100 100
Let =1 or =1
10 + j 100
2
( 2
) 100 4 + 2 (100 2 )2
1/ 2
(100 )
2
Thus 100 + = 10, 000 100 + 10, 000 10, 000 = 0
4 2 2 6 4 2
Therealsolutionfor are = 1rad/sec.
848
100 2
1
G1 ( j1) = tan = 264.23
o

10 =1
Equating Td =
( 264.23 o
180
o
)
=1
180
84.23
= = 1.47 rad
180
Thusthemaximumtimedelayforstability
is
Td = 1.47 sec.
(b) Td = 1 sec.
s j
100 Ke 100 Ke
G( s) = G ( j ) =
(
s s + 10 s + 100
2
) (
10 + j 100
2 2
)
Attheintersectonthenegativerealaxis, = 1.42 rad/sec.
G ( j1.42) = 0.7107 K .
Thesystemisstablefor
0<K<1.407
849
830(a)K=0.1
Td s
10e Td s
G ( s ) = = G1 ( s )e
(
s s + 10 s + 100
2
)
10 10
Let =1 or = 1
10 + j 100
2
( 2
) 100 4 + 2 (100 2 )2
1/ 2
100 + 10,000 100 = 0 Therealsolutionsfor is = 0.1rad/sec.

6 4 2
Thus
100 2 1
G1 ( j 0.1) = tan = 269.43
o

10 = 0.1
Equate Td =
( 269.43 o
180
o
) = 1.56 radWehave Td = 15.6 sec.
= 0.1
180
o
Wehavethemaximumtimedelay
forstabilityis15.6sec.
830(b) Td = 0.1 sec.
0.1 s 0.1 j
100 Ke 100 Ke
G( s) = G ( j ) =
(
s s + 10 s + 100
2
) (
10 + j 100
2 2
)
Attheintersectonthenegativerealaxis,
= 6.76 rad/sec. G ( j 6.76 ) = 0.1706 K
850
Thesystemisstablefor
0<K<5.86
8-31)
MATLAB code:
s = tf('s')
%a)
figure(1);
K=1
num_GH_a= K;
den_GH_a=s*(s+1)*(s+1);
nyquist(GH_a)
%b)
figure(2);
K=20
num_GH_b= K;
den_GH_b=s*(s+1)*(s+1);
nyquist(GH_b)
sisotool;
851
By using sisotool and importing the loop transfer function, different values of K has been tested which
resulted in a stable system when K<2, and unstable system for K>2. Following diagrams correspond to
margin of stability:
3 30
2 20
10
1
0
0
-10
-1
-20
-2
-30
G.M.: -0.282 dB
-3 Freq: 1 rad/sec
-4 -3 -2 -1 0 1 -40
Unstable loop
100
-90
0
-135
-100
-200 -180
0
-180
-225
-360 P.M.: -0.925 deg
Freq: 1.02 rad/sec
-540 -270
-1 0 1 2 0
10 10 10 10 10
Part(a): small K resulted in stable system as shown below for K=0.219:
852
3 10
2 0
-10
1
-20
0
-30
-1
-40
-2
-50
G.M.: 19.2 dB
-3 Freq: 1 rad/sec
-4 -3 -2 -1 0 1 -60
Stable loop
0
-90
-50
-135
-100
-150 -180
0
-225
-180
P.M.: 66.3 deg
Freq: 0.21 rad/sec
-360 -270
-2 -1 0 1 2 0
10 10 10 10 10 10
Nyquist Diagram
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-1 -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0
Real Axis
Part(b): Large K resulted in unstable system as shown below for K=20:
853
3 50
40
2
30
1
20
0 10
-1 0
-10
-2
-20
G.M.: -20 dB
-3 Freq: 1 rad/sec
-4 -3 -2 -1 0 1 -30
Unstable loop
-90
0
-50 -135
-100 -180
0
-90
-180 -225
-270 P.M.: -47.8 deg
-450 -270
-1 0 1 2 0
10 10 10 10 10
Nyquist Diagram
30
20
10
Imaginary Axis
-10
-20
-30
-15 -10 -5 0 5
Real Axis
854
The system is stable for small value of K, since there is no encirclement of the s = -1
The system is unstable for large value of K, since the locus encirclement the twice in
CCW; which means two poles are in the right half s-plane.
8-32)(a)Thetransferfunction(gain)forthesensoramplifiercombinationis10V/0.1in=100V/in.Thevelocity
offlowofthesolutionis
3
10 in / sec
v= = 100 in/sec
0.1 in
ThetimedelaybetweenthevalveandthesensorisTd = D / v sec.Thelooptransferfunctionis
Td s
100 Ke
G( s) = 2

s + 10 s + 100
855
856
8-33)
(a)Thetransferfunction(gain)forthesensoramplifiercombinationis1V/0.1in=10V/in.Thevelocity
offlowofthesolutionsis
3
10 in / sec
v= = 100 in / sec
0.1 in
Thetimedelaybetweenthevalveandsensoris Td = D / v sec.Thelooptransferfunctionis
Td s
10 Ke
G( s) = 2

s + 10 s + 100
(b)K=10:
j Td
100e
G ( j ) =
Td s
G ( s ) = G1 ( s )e
(100 ) + j102
100
(100 )
2
=1 + 100 = 10, 000
2 2
Setting
( 100
2
) + j10
4 2
Thus, 100 = 0 Real solutions: = 0, = 10 rad / sec
G1 ( j10) = tan
1 10 = 90
2
o
100 =10
Thus,
90
o
10Td = o
= rad
180 2
Thus,

Td = = 0.157 sec
20

Maximum D = vTd = 100 0.157 = 15.7 in
857
(c)D=10in.
0.1s
D 10 10 Ke
Td = = = 0.1 sec G( s) = 2

v 100 s + 10 s + 100
TheNyquistplotofG ( j ) intersectsthenegativerealaxisat = 12.09 rad/sec.G ( j ) = 0.0773K
Forstability,the
maximumvalueofKis12.94.
8-34)
The system (GH) has zero poles in the right of s plane: P=0.
According to Nyquist criteria (Z=N+P), to ensure the stability which means the number of right poles of
1+ GH=0 should be zero (Z=0), we need N clockwise encirclements of Nyquist diagram about -1+0j
point. That is N=-P or in other words, we need P counter-clockwise encirclement about -1+0j. In this
case, we need 0 CCW encirclements.
8-34(a) According to Nyquist diagrams, this happens when K<-1. The three Nyquist diagrams are
plotted with K=-10, K=-1, K=10 as examples:
MATLAB code:
s = tf('s')
%a)
figure(1);
K=-10
num_G_a = K ;
858
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
H_a = num_H_a/den_H_a;
OL_a = G_a*H_a
nyquist(OL_a)
Case 1) Nyquist graph, K=-10: margin of stability K<-1 unstable
figure(2);
K=-1
num_G_a = K ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
%CL_a = G_a/(1 + G_a*H_a);
OL_a = G_a*H_a
nyquist(OL_a)
859
Case 2) Nyquist graph, K=-1: marginally unstable
figure(3);
K=10
num_G_a = K ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
%CL_a = G_a/(1 + G_a*H_a);
OL_a = G_a*H_a
nyquist(OL_a)
860
Case 3) Nyquist graph, K=10: stable case, -1<K no CCW encirclement about -1+0j point
861
8-34 (b)
For K<-1 (unstable), there will be 1 real pole in the right hand side of s-plane for the closed loop system,
by running the following code.
K=-10
num_G_a = K ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
OL_a = G_a*H_a
nyquist(OL_a)
CL=1/(1+OL_a)
pole(CL)
K =
-10
Transfer function:
-10 s - 20
---------------------
s^3 + 3 s^2 + 4 s + 2
Transfer function:
s^3 + 3 s^2 + 4 s + 2
----------------------
s^3 + 3 s^2 - 6 s - 18
ans =
2.4495
-3.0000
-2.4495
For K=-1 (marginally unstable), there will be 2 negative complex conjugate poles and a pole at zero for
the closed loop system, by running the following code.
K=-1
num_G_a = K ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
862
OL_a = G_a*H_a
nyquist(OL_a)
CL=1/(1+OL_a)
pole(CL)
K=
-1
Transfer function:
-s - 2
---------------------
s^3 + 3 s^2 + 4 s + 2
Transfer function:
s^3 + 3 s^2 + 4 s + 2
---------------------
s^3 + 3 s^2 + 3 s
ans =
0
-1.5000 + 0.8660i
-1.5000 - 0.8660i
Note: you may also wish to use MATLAB sisotool.

See alternative solution to 8-38.
863
8-34(c) The Characteristic Equation is: s3 + 3 s2 + (4+K) s + 2+2K

Using Routh criterions, the coefficient table is as follows:
s3 1 4+K
s2 3 2K+2
s1 K+10 0
s0 2K+2 0
The system is stable if the content of the 1st column is positive:

K+10>0 K>-10
2K+2>0 K>-1
which is consistent with the results of the Nyquist diagrams. For K>-1 system is stable.
8-35) (a) M r = 2.06, r = 9.33 rad / sec, BW = 15.2 rad / sec
(b)
1 4 2
M r = = 2.06 + 0.0589 = 0 Thesolutionfor < 0.707 is = 0.25.
2
2 1
2 9.33
r 1 2 = 9.33 rad / sec Thus n = = 9.974 rad / sec
0.9354
n
2
99.48 19.94
GL ( s ) = = = BW=15.21rad/sec
s ( s + 2 n ) s ( s + 4.987) s (1 + 0.2005 s )
864
8-36) Assuming a unity feedback loop (H=1), G(s)H(s)=G(s)

(a)
5
G ( s ) =
s(1 + 0.5s )(1 + 0.1s )
865
836(b)
10
G ( s ) =
s(1 + 0.5s )(1 + 0.1s )
(c)
500
G ( s ) =
( s + 1.2)( s + 4 )( s + 10 )
866
(d)
10( s + 1)
G ( s ) =
s( s + 2 )( s + 10 )
836(e)
0.5
G ( s ) =
(
s s + s +1
2
)
867
(f)
s
100e
G ( s ) =
(
s s + 10 s + 50
2
)
868
(g)
s
100e
G ( s ) =
(
s s + 10 s + 100
2
)
836(h)
10( s + 5)
G ( s ) =
(
s s + 5s + 5
2
)
(a)
869
MATLAB code:
s = tf('s')
num_G_a= 5;
den_G_a=s*(0.5*s+1)*(0.1*s+1);
G_a=num_G_a/den_G_a
margin(G_a)
Bode diagram:
(b)
MATLAB code:
s = tf('s')
num_G_a= 10;
den_G_a=s*(1+0.5*s)*(1+0.1*s);
G_a=num_G_a/den_G_a
margin(G_a)
870
(c)
MATLAB code:
s = tf('s')
num_G_a= 500;
den_G_a=s*(s+1.2)*(s+4)*(s+10);
G_a=num_G_a/den_G_a
margin(G_a)
871
(d)
MATLAB code:
s = tf('s')
num_G_a= 10*(s+1);
den_G_a=s*(s+2)*(s+10);
G_a=num_G_a/den_G_a
margin(G_a)
872
(e)
MATLAB code:
s = tf('s')
num_G_a= 0.5;
den_G_a=s*(s^2+s+1);
G_a=num_G_a/den_G_a
margin(G_a)
873
(f)
MATLAB code:
s = tf('s')
num_G_a= 100*exp(-s);
den_G_a=s*(s^2+10*s+50);
G_a=num_G_a/den_G_a
margin(G_a)
874
(g)
MATLAB code:
s = tf('s')
num_G_a= 100*exp(-s);
den_G_a=s*(s^2+10*s+100);
G_a=num_G_a/den_G_a
margin(G_a)
875
(h)
MATLAB code:
s = tf('s')
num_G_a= 10*(s+5);
den_G_a=s*(s^2+5*s+5);
G_a=num_G_a/den_G_a
margin(G_a)
876
877
8-37)
MATLAB code:
s = tf('s')
num_GH_a= 25*(s+1);
den_GH_a=s*(s+2)*(s^2+2*s+16);
GH_a=num_GH_a/den_GH_a
margin(GH_a)
878
8-38) MATLAB code:

s = tf('s')
num_G_a= 25*(s+1);
den_G_a=s*(s+2)*(s^2+2*s+16);
G_a=num_G_a/den_G_a
margin(G_a)
Bode diagram: PM=101 deg, GM=3.91 dB @ 4.22 rad/sec
Bode Diagram
Gm = 3.91 dB (at 4.22 rad/sec) , Pm = 101 deg (at 1.08 rad/sec)
20
0
Magnitude (dB)
-20
-40
-60
-80
-100
-45
-90
Phase (deg)
-135
-180
-225
-270
-1 0 1 2
10 10 10 10
Frequency (rad/sec)
8-38 Alternative solution

MATLABcode:
s = tf('s')
%a)
figure(1);
num_G_a = 1 ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
879
CL_a = G_a/(1 + G_a*H_a);
margin(CL_a)
sisotool
Bodediagram:fork=1
Bode Diagram
Gm = Inf , Pm = Inf
0
-10
Magnitude (dB)
-20
-30
-40
0
Phase (deg)
-45
-90
-1 0 1 2
10 10 10 10
Frequency (rad/sec)

UsingMATLABsisotool,thetransferfunctiongaincanbeiterativelychangedinordertoobtaindifferentphase
margins. By changing the gain K between very small and very big numbers, it was found that the closed loop
systemarestable(positivePM)foreverypositiveKinthissystem.
K=0.034
880
10 -20
5 -40
-60
0
-80
-5
-100 G.M.: Inf

-10 Freq: Inf
-2 -1.5 -1 -0.5 0
Stable loop
0
0
-50 -45
-100 -90
0
-135
-45
P.M.: Inf
Freq: NaN
-90 -180
-2 -1 0 1 2 -1 0 1 2
10 10 10 10 10 10 10 10 10
K=59.9
10 40
5 20
0
0
-20
-5
-40 G.M.: Inf

-10 Freq: Inf
-2 -1.5 -1 -0.5 0
Stable loop
40
0
20
-45
0
-20 -90
90
-135
0
P.M.: 7.67 deg
Freq: 7.89 rad/sec
-90 -180
-1 0 1 2 -1 0 1 2
10 10 10 10 10 10 10 10
881
InordertotestthenegativerangeofK,1wasmultipliedtothelooptransferfunctionthroughthefollowing
code,andsisotoolwasusedagain.
figure(1);
num_G_a = -1 ;
den_G_a =(s+1);
num_H_a = (s+2);
den_H_a = (s^2+2*s+2);
CL_a = G_a/(1 + G_a*H_a);
margin(CL_a)
sisotool
atK=1,marginofstabilityisobservedasPM~=0:
K=1
1 10
0.5 -10
-20
0 -30
-40
-0.5
-50
-60
G.M.: -0.0609 dB
-1 Freq: 0 rad/sec
-4 -2 0 2 4 -70
Unstable loop
40 180
20
0 135
-20
-40 90
135
90
45
45 P.M.: -11.8 deg
Freq: 0.137 rad/sec
0 0
-4 -2 0 2 -1 0 1 2
10 10 10 10 10 10 10 10
ThesystemisstableforK>1asfollows:K=0.6
882
1 0
0.5 -20
-40
0
-60
-0.5
-80 G.M.: 4.37 dB

-1 Freq: 0 rad/sec
-6 -4 -2 0 2 4
Stable loop
50
180
0 135
-50 90
180
45
90
P.M.: Inf
Freq: NaN
0 0
-2 -1 0 1 2 -1 0 1 2
10 10 10 10 10 10 10 10 10

AndthesystemisunstableforK<1:K=3
1 20
0.5 0
-20
0
-40
-0.5
-60 G.M.: -9.59 dB

-1 Freq: 0 rad/sec
-6 -4 -2 0 2 4
Unstable loop
20
180
0
135
-20
-40 90
180
45
90
P.M.: -129 deg
Freq: 1.84 rad/sec
0 0
-2 -1 0 1 2 -1 0 1 2
10 10 10 10 10 10 10 10 10

*CombiningtheindividualrangesforK,thesystemwillbestableintherangeofK>1
883
839SeesampleMATLABcodeinParte.TheMATLABcodesareidenticaltoproblem836.
(a)
K
G ( s ) =
s(1 + 0.1s )(1 + 0.5s )
TheBodeplotisdonewithK=1.
GM=21.58dBForGM=20dB,
Kmustbereducedby1.58dB.
ThusK=0.8337
o o
PM= 60.42 .ForPM= 45
Kshouldbeincreasedby5.6dB.
Or,K=1.91
(b)
K ( s + 1)
G( s) = The
s(1 + 0.1s )(1 + 0.2 s)(1 + 0.5s )
BodeplotisdonewithK=1.
GM=19.98dB.ForGM=20dB,
K 1.
o o
PM= 86.9 .ForPM= 45

Kshouldbeincreasedby8.9dB.
Or,K=2.79.
884
839 (c)Seethetopplot
K
G ( s ) = 3

( s + 3)
GM=46.69dB
PM=infinity.
ForGM=20dBKcanbe
increasedby26.69dBorK=21.6.
ForPM=45deg.Kcanbe
increasedby28.71dB,or
K=27.26.
(d)Seethemiddleplot
K
G( s) = 4

( s + 3)
GM=50.21dB
PM=infinity.
ForGM=20dBKcanbe
increasedby30.21dBorK=32.4
ForPM=45deg.Kcanbe
increasedby38.24dB,or
K=81.66
(e)Seethebottomplot

885
s
Ke MATLAB code:
G ( s ) =
(
s 1 + 0.1s + 0.01s
2
) s = tf('s')
num_G_a= exp(-s);
GM=2.97dB;PM=26.58deg
den_G_a=s*(0.01*s^2+0.1*s+1);
ForGM=20dBKmustbe G_a=num_G_a/den_G_a
margin(G_a)
decreasedby17.03dBor
K=0.141.
ForPM=45deg.Kmustbe
decreasedby2.92dBorK=0.71.
886
839(f)
K (1 + 0.5s )
G ( s ) =
(
s s + s +1
2
)
GM=6.26dB
PM=22.24deg
ForGM=20dBKmustbedecreasedby13.74dBor
K=0.2055.
ForPM=45degKmustbedecreasedby3.55dBor
K=0.665.
840(a)
10 K
G ( s ) =
s(1 + 0.1s )(1 + 0.5s )
Thegainphaseplotisdonewith
K=1.
GM=1.58dB
PM=3.95deg.

ForGM=10dB,Kmustbedecreasedby8.42
dBorK=0.38.
ForPM=45deg,Kmustbedecreasedby14dB,orK=0.2.
For M r = 1.2 ,Kmustbedecreasedto0.16.
887
SampleMATLAB code:
s = tf('s')
num_G_a= 10;
den_G_a=s*(1+0.1*s)*(0.5*s+1);
G_a=num_G_a/den_G_a
nichols(G_a)
(b)
888
5K ( s + 1)
G( s) =
s(1 + 0.1s )(1 + 0.2 s)(1 + 0.5s )
TheGainphaseplotisdonewith
K=1.
GM=6dB
PM=22.31deg.
ForGM=10dB,Kmustbedecreasedby4

dBorK=0.631.
ForPM=45deg,Kmustbe
decreaseby5dB.
For M r = 1.2 ,Kmustbedecreasedto0.48.
840(c)
10 K
G ( s ) =
(
s 1 + 0.1s + 0.01s
2
)
Thegainphaseplotisdonefor
K=1.
GM=0dB M r =
PM=0deg
ForGM=10dB,Kmustbedecreasedby10
dBorK=0.316.
ForPM=45deg,Kmustbedecreasedby5.3
dB,or
K=0.543.
For M r = 1.2, K mustbedecreasedto0.2213.
(d)
889
s
Ke
G ( s ) =
(
s 1 + 0.1s + 0.01s
2
)
Thegainphaseplotisdonefor
K=1.
GM=2.97dB M r = 3.09
PM=26.58deg
ForGM=10dB,Kmustbedecreasedby7.03
dB,K=0.445.

ForPM=45deg,Kmustbedecreasedby
2.92dB,or
K=0.71.
For M r = 1.2, K = 0.61.
890
8-41
MATLAB code:
s = tf('s')
%a)
num_GH_a= 1*(s+1)*(s+2);
den_GH_a=s^2*(s+3)*(s^2+2*s+25);
CL_a = GH_a/(1+GH_a)
figure(1);
bode(CL_a)
%b)
figure(2);
rlocus(GH_a)
%c)
num_GH_c= 53*(s+1)*(s+2);
den_GH_c=s^2*(s+3)*(s^2+2*s+25);
GH_c=num_GH_c/den_GH_c;
figure(3);
nyquist(GH_c)
xlim([-2 1])
ylim([-1.5 1.5])
%d)
num_GH_d= (s+1)*(s+2);
den_GH_d=s^2*(s+3)*(s^2+2*s+25);
GH_d=num_GH_d/den_GH_d;
CL_d = GH_d/(1+GH_d)
figure(4);
margin(CL_d)
sisotool
Part (a), Bode diagram:
891
Bode Diagram
50
0
Magnitude (dB)
-50
-100
-150
0
-90
Phase (deg)
-180
-270
-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/sec)
Part (b), Root locus diagram:
Root Locus
30
20
10
Imaginary Axis
-10
-20
-30
-40 -30 -20 -10 0 10 20
Real Axis
Part (c), Gain and frequency that instability occurs: Gain=53, Freq = 4.98 rad/sec, as seen in the data
point in the figure:
892
Nyquist Diagram
1.5
System: GH_a
0.5
Real: -1
Imag: 0.0174
Imaginary Axis
0
-0.5
-1
-1.5
-2 -1.5 -1 -0.5 0 0.5 1
Real Axis
Part (d), Gain and

frequency that
instability occurs:
Gain=0.127 at PM
=20 deg:
By running sisotool
command in
MATLAB, the
transfer functions
are imported and the
gain is iteratively
changed until the
phase margin of
PM=20 deg is
achieved. The Part (e): The corresponding gain margin is GM = 22.7 dB is seen in the
corresponding Gain figure
is K= 0.127.
893
842(a)Gaincrossoverfrequency=2.09rad/sec PM=115.85deg
Phasecrossoverfrequency=20.31rad/sec GM=21.13dB
(b)Gaincrossoverfrequency=6.63rad/sec PM=72.08deg
(c)Gaincrossoverfrequency=19.1rad/sec PM=4.07deg
(d)ForGM=40dB,reducegainby(4021.13)dB=18.7dB,orgain=0.116 nominalvalue.
(e)ForPM=45deg,themagntudecurvereads10dB.Thismeansthattheloopgaincanbeincreasedby
10dBfromthenominalvalue.Orgain=3.16 nominalvalue.
(f)Thesystemistype1,sincetheslopeof G ( j ) is20dB/decadeas 0.
(g)GM=12.7dB.PM=109.85deg.
(h)Thegaincrossoverfrequencyis2.09rad/sec.Thephasemarginis115.85deg.
Set
115.85
o
Td = 2.09Td = o
= 2.022 rad
180
Thus,themaximumtimedelayis Td = 0.9674 sec.
894
843(a)Thegainisincreasedtofourtimesitsnominalvalue.Themagnitudecurveisraisedby12.04dB.
Gaincrossoverfrequency=10rad/sec PM=46deg
(b)TheGMthatcorrespondstothenominalgainis21.13dB.TochangetheGMto20dBweneedto
increasethegainby1.13dB,or1.139timesthenominalgain.
(c)TheGMis21.13dB.Theforwardpathgainforstabilityis21.13dB,or11.39.
(d)ThePMforthenominalgainis115.85deg.ForPM=60deg,thegaincrossoverfrequencymustbe
movedtoapproximately8.5rad/sec,atwhichpointthegainis10dB.Thus,thegainmustbeincreased
by10dB,orbyafactorof3.162.
(e)Withthegainattwiceitsnominalvalue,thesystemisstable.Sincethesystemistype1,thesteadystate
errorduetoastepinputis0.
(f)Withthegainat20timesitsnominalvalue,thesystemisunstable.Thusthesteadystateerrorwouldbe
infinite.
(g)Withapuretimedelayof0.1sec,themagnitudecurveisnotchanged,butthethephasecurveissubject
toanegativephaseof 0.1 rad.ThePMis
PM = 115.85 0.1 gain crossover frequency = 115.85 0.209 = 115.64 deg
Thenewphasecrossoverfrequencyisapproximately9rad/sec,wheretheoriginalphasecurveis
reducedby0.9rador51.5deg.Themagnitudeofthegaincurveatthisfrequencyis10dB.
Thus,thegainmarginis10dB.
(h)Whenthegainissetat10timesitsnominalvalue,themagnitudecurveisraisedby20dB.Thenew
gaincrossoverfrequencyisapproximately17rad/sec.Thephaseatthisfrequencyis30deg.
Thus,setting
30
o
Td = 17Td = o
= 0.5236 Thus Td = 0.0308 sec.
180
895
8-44
MATLAB code:
s = tf('s')
%using pade cammand for PADE approximation of exponential term
num_G_a= pade((80*exp(-0.1*s)),2);
den_G_a=s*(s+4)*(s+10);
CL_a = G_a/(1+G_a)
OL_a = G_a*1;
%(a)
figure(1)
nyquist(OL_a)
%(b) and (c)

figure(2);
margin(CL_a)
Part (a), Nyquist diagram:
Nyquist Diagram
8
2
Imaginary Axis
-2
-4
-6
-8
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4
Real Axis
Part (b) and (c), Bode diagram:

Using Margin command, the gain and phase margins are obtained as GM = 6.3 dB, PM = 42.6 deg:
896
Bode Diagram
Gm = 6.3 dB (at 4.02 rad/sec) , Pm = 42.6 deg (at 2.9 rad/sec)
50
0
Magnitude (dB)
-50
-100
-150
360
180
Phase (deg)
-180
-360
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
897
845(a)BodePlot:
Forstability:166(44.4dB)<K<7079(77dB) Phasecrossoverfrequencies:7rad/secand85rad/sec
898
NyquistPlot:
899
845(b)RootLoci.
8100
846(a)NyquistPlot
BodePlot
8101
RootLocus
8102
8-47
MATLAB code:
%a)
k=1
num_GH= k*(s+1)*(s+5);
den_GH=s*(s+0.1)*(s+8)*(s+20)*(s+50);
GH=num_GH/den_GH;
CL = GH/(1+GH)
figure(1);
bode(CL)
figure(2);
OL = GH;
nyquist(GH)
xlim([-1.5 0.5]);
ylim([-1 1]);
sisotool
Part (a), Bode diagram:
8103
Bode Diagram
0
-50
Magnitude (dB)
-100
-150
-200
0
-90
Phase (deg)
-180
-270
-4 -3 -2 -1 0 1 2 3
10 10 10 10 10 10 10 10
Frequency (rad/sec)
Part (a), Nyquist diagram:
Nyquist Diagram
1
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1
-1.5 -1 -0.5 0 0.5
Real Axis
Part (a), range of K for stability: By running sisotool command in MATLAB, the transfer functions
are imported and the gain is iteratively changed until the phase margin of PM=0 deg is achieved (where
K = ) which is the margin of stability. The stable rang for K is K>8.16x104:
8104
Part (b), Root-locus diagram, K and at the points where the root loci cross the j-axis:
As can be seen in the figure at K=8.13x104 and =33.8 rad/sec, the poles cross the j axis. Both of
these values are consistent with the results of part(a) from sisotool.
8105
Root Locus
150
System: CL
100 Gain: 8.13e+004
Pole: 0.0788 + 33.8i
Damping: -0.00233
Overshoot (%): 101
50 Frequency (rad/sec): 33.8
Imaginary Axis
-50 System: CL
Gain: 7.99e+004
Pole: -0.045 - 33.5i
Damping: 0.00134
-100 Overshoot (%): 99.6
-150
-200 -150 -100 -50 0 50 100
Real Axis
8106
848BodeDiagram
WhenK=1,GM=68.75dB,PM=90deg.ThecriticalvalueofKforstabilityis2738.
8107
849(a)Forwardpathtransferfunction:
L ( s) K a Ki ( Bs + K )
G( s) = = K aG p ( s ) =
E ( s) o
where
(
o = 0.12 s ( s + 0.0325 ) s + 2.5675 s + 6667
2
)

(
= s 0.12 s + 0.312 s + 80.05 s + 26
3 2
)
43.33( s + 500)
G ( s ) =
(
s s + 2.6 s + 667.12 s + 216.67
3 2
)
(b)BodeDiagram:
8108
Gaincrossoverfrequency=5.85rad/sec PM=2.65deg.
8109
849(c)ClosedloopFrequencyResponse:
M r = 17.72, r = 5.75 rad / sec, BW = 9.53 rad / sec
8110
8-50
(a)
(b)
(c) to (e)
MATLAB code:
s = tf('s')
m = 0.11;
r = 0.015;
d = 0.03;
g = 9.8;
L = 1.0;
J = 9.99*10^-6
K=1
num_GH= K*m*g*d;
den_GH=L*(J/r^2+m)*s^2;
GH=num_GH/den_GH;
CL = GH/(1+GH)
%c)
figure(1);
nyquist(GH)
xlim([-1.5 0.5]);
ylim([-1 1]);
%d)
figure(2);
margin(CL)
Part (c): since the system is a double integrator (1/s2), the phase is always -180 deg, and the system is
always marginally stable for any K, leading to a complicated control problem.
8111
Nyquist Diagram
1
0.8
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-1
-1.5 -1 -0.5 0 0.5
Real Axis
Part (d), Bode diagram:

As explained in section (c), since the system is always marginally stable, GM=inf and PM = 0, as can be
seen by MATLAB MARGIN command, resulting in the following figure:
8112
Bode Diagram
Gm = Inf dB (at Inf rad/sec) , Pm = 0 deg (at 0.647 rad/sec)
150
System: CL
100 Frequency (rad/sec): 0.458
Magnitude (dB): 143
Magnitude (dB)
50
0
System: CL
-100
-180
-225
Phase (deg)
-270
-315
-360
-2 -1 0 1
10 10 10 10
Frequency (rad/sec)
Part (e), Mr,= 143 dB, r = 0.458 rad/sec, and BW= 0.712 rad/sec as can be seen in the data points in
the above figure.
8-51 (a)WhenK=1,thegaincrossoverfrequencyis8rad/sec.
(b)WhenK=1,thephasecrossoverfrequencyis20rad/sec.
(c)WhenK=1,GM=10dB.
(d)WhenK=1,PM=57deg.
(e)WhenK=1, M r = 1.2.
(f)WhenK=1, r = 3 rad/sec.
(g)WhenK=1,BW=15rad/sec.
(h)WhenK=10dB(0.316),GM=20dB
(i)WhenK=10dB(3.16),thesystemismarginallystable.Thefrequencyofoscillationis20rad/sec.
(j)Thesystemistype1,sincethegainphaseplotof G ( j ) approachesinfinityat90deg.Thus,the
steadystateerrorduetoaunitstepinputiszero.
8113
8-52WhenK=5dB,thegainphaseplotof G ( j ) israisedby5dB.
(a)Thegaincrossoverfrequencyis~10rad/sec.
(b)Thephasecrossoverfrequencyis~20rad/sec.
(c)GM=5dB.
(d)PM=~34.5deg.
(e)WhenK=5,Mr=~2(smallestcircletangenttoanMcircle).
(f) r = 15 rad/sec
(g)BW=30rad/sec
(h)WhenK=30dB,theGMis40dB(shiftthegraphofK=1,30dbsdown).
WhenK=10dB,thegainphaseplotofG ( j ) israisedby10dB.
(a)Thegaincrossoverfrequencyis20rad/sec.
(b)Thephasecrossoverfrequencyis20rad/sec.
(c)GM=0dB.
(d)PM=0deg.
(e)WhenK=10,Mr=~1.1(smallestcircletangenttoanMcircle).
(f) r = 5 rad/sec
(g)BW=~40rad/sec
(h)WhenK=30dB,theGMis40dB(shiftthegraphofK=1,30dbsdown).
8-53
Since the function has exponential term, PADE command has been used to obtain the transfer function.
80e 0.1s
G ( s) H ( s) =
s ( s + 4)( s + 10)
MATLAB code:
s = tf('s')
%a)
8114
num_GH= pade(80*exp(-0.1*s),2);
den_GH=s*(s+4)*(s+10);
GH=num_GH/den_GH;
CL = GH/(1+GH)
BW = bandwidth(CL)
bode(CL)
%b)
figure(2);
nichols(GH)
Part(a), Nicholas diagram:
Nichols Chart
40
20
-20
Open-Loop Gain (dB)
-40
-60
-80
-100
-120
-140
-270 -180 -90 0 90 180 270
Open-Loop Phase (deg)
Part(b), Bode diagram:
8115
Bode Diagram
50
0
Magnitude (dB)
-50
-100
-150
360
180
Phase (deg)
-180
-360
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
G
8- 54) Note: GCL =
1+ G
To draw the Bode and polar plots use the closed loop transfer function, GCL, and find BW. Use G to
obtain the gain-phase plots and Gm and Pm. Use the Bode plot to graphically obtain Mr.
Sample MATLAB code:
s = tf('s')
%a)
num_G= 1+0.1*s;
den_G=s*(s+1)*(0.01*s+1);
G=num_G/den_G
figure(1)
nyquist(G)
figure(2)
margin(G)
GCL = G/(1+G)
BW = bandwidth(GCL)
figure(3)
bode(GCL)
8116
Transfer function:
0.1 s + 1
-----------------------
0.01 s^3 + 1.01 s^2 + s
Transfer function:
0.001 s^4 + 0.111 s^3 + 1.11 s^2 + s
--------------------------------------------------------------
0.0001 s^6 + 0.0202 s^5 + 1.041 s^4 + 2.131 s^3 + 2.11 s^2 + s
BW =
1.2235
8117
8118
8-55) (a)ThephasemarginwithK=1andTd = 0 secisapproximately57deg.ForaPMof40deg,thetimedelay
producesaphaselagof17deg.Thegaincrossoverfrequencyis8rad/sec.
Thus,
17
o
Td = 17 = Thus = 8 rad / sec
o
o
= 0.2967 rad / sec
180
0.2967
Td = = 0.0371sec
8
(b)WithK=1,formarginalstability,thetimedelaymustproduceaphaselagof57deg.
Thus,at = 8 rad/sec,
57
o
0.9948
Td = 57 =
o
o
= 0.9948 rad Td = = 0.1244 sec
180 8
8119
856(a)ThephasemarginwithK=5dBandTd = 0 isapproximately34.5deg.ForaPMof30deg,thetime
delaymustproduceaphaselagof4.5deg.Thegaincrossoverfrequencyis10rad/sec.Thus,
4.5
o
0.0785
Td = 4.5 =
o
o
= 0.0785 rad Thus Td = = 0.00785 sec
180 10
(b)WithK=5dB,formarginalstability,thetimedelaymustproduceaphaselagof34.5deg.
Thusat = 10 rad/sec,
34.5
o
0.602
Td = 34.5 =
o
o
= 0.602 rad Thus Td = = 0.0602 sec
180 10
857)ForaGMof5dB,thetimedelaymustproduceaphaselagof34.5degat = 10 rad/sec.Thus,
34.5
o
0.602
Td = 34.5 =
o
o
= 0.602 rad Thus Td = = 0.0602 sec
180 10
2 s
Y ( s) e
G( s) = =
E ( s) (1 + 10 s )(1 + 25s )
FromtheBodediagram,phasecrossoverfrequency=0.21rad/sec GM=21.55dB
8120
gaincrossoverfrequency=0rad/sec PM=infinite
(b)
1
G(s) =
(1 + 10 s )(1 + 25s ) (1 + 2 s + 2 s 2
)
FromtheBodediagram,phasecrossoverfrequency=0.26rad/sec GM=25dB
(c)
1 s
G(s) =
(1 + s )(1 + 10 s )(1 + 2 s )
Sample MATLAB code

s = tf('s')
%a)
num_G=exp(-2*s);
den_G=(10*s+1)*(25*s+1);
G=num_G/den_G
figure(1)
margin(G)
8121
858(continued)Bodediagramsforallthreeparts.
8122
s
e
G( s) =
(1 + 10 s )(1 + 25s )
(b)
1
G( s) =
(1 + 10s )(1 + 25s ) (1 + s + 0.5s 2
)
FromtheBodediagram,phasecrossoverfrequency=0.367rad/secGM=30.72dB
(c)
(1 0.5s )
G( s) =
(1 + 10 s )(1 + 25s )(1 + 0.5 s )
FromtheBodediagram,phasecrossoverfrequency=0.3731rad/secGM=31.18dB
8123
Plots859(ac)
8124
860SensitivityPlot:
8125
max = 5.75 rad/sec

M
SG = 17.15
max
8-61)
(a)
(b)
(c)&(d)
MATLAB code:
s = tf('s')
%c)
K = 1
num_GH= K*(1.151*s+0.1774);
den_GH=(s^3+0.739*s^2+0.921*s);
GH=num_GH/den_GH;
CL = GH/(1+GH)
sisotool
%(d)
figure(1)
margin(CL)
Part (c), range of K for stability:

Sisotool Result shows that by changing K between 0 and inf., all the roots of closed loop system remain
in the left hand side plane and PM remains positive. Therefore, the system is stable for all positive K.
8126
4
Root
x 10 Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
3 200
2
180
1
160
0
-1
140
-2
120 G.M.: Inf
-3 Freq: Inf
-0.4 -0.3 -0.2 -0.1 0
Stable loop
Bode Editor for Closed Loop 1 (CL1) 100
100
0
0 -45
-100 -90
0
-135
-90
P.M.: 0.00149 deg
Freq: 2.26e+004 rad/sec
-180 -180
-2 0 2 4 6 -2 -1 0 1 2
10 10 10 10 10 10 10 10 10 10
Part (d), Bode, GM & PM for K=1:
8127
Bode Diagram
Gm = Inf dB (at Inf rad/sec) , Pm = 61.3 deg (at 1.59 rad/sec)
20
Magnitude (dB) 0
-20
-40
-60
-80
0
-45
Phase (deg)
-90
-135
-180
-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/sec)
8128
Chapter 9
91ForwardpathTransferFunction:
M ( s) K
G( s) = = 3 2

1 M ( s) s + ( 20 + a ) s + ( 200 + 20a ) s + 200a K
Ramperrorconstant:
K 200 a
K v = lim sG ( s ) = = = 5 Thusa=10 K=2000
s 0 200 + 20a 200 + 20 a
G ( s ) =
2000
Gc ( s ) =
G(s)
=
(
20 s + 10 s + 100
2
)
(
s s + 30 s + 400
2
) Gp (s) (s 2
+ 30 s + 400 )
Themaximumovershootoftheunitstepresponseis0percent.
M ( s) K
G( s) = = 3 2

1 M ( s) s + ( 20 + a ) s + ( 200 + 20a ) s + 200a K
Ramperrorconstant:
K 200 a
K v = lim sG ( s ) = = = 9 Thusa=90 K=18000
s 0 200 + 20a 200 + 20 a
91
G ( s ) =
18000
Gc ( s ) =
G(s)
=
(
180 s + 10 s + 100
2
)
(
s s + 110 s + 2000
2
) G p ( s) ( s + 110s + 2000 )
2
Themaximumovershootoftheunitstepresponseis4.3percent.
Fromtheexpressionfortheramperrorconstant,weseethatasaorKgoestoinfinity, Kv approaches10.
Thusthemaximumvalueof Kv thatcanberealizedis10.ThedifficultieswithverylargevaluesofKand
aarethatahighgainamplifierisneededandunrealisticcircuitparametersareneededforthecontroller.
93)Thecloselooptransferfunctionis:
Comparingwithsecondordersystem:
and2
exp 0.254 =0.4
3 n=1.14
=1.09
K=1.42
9-4) The forward path transfer function of the system is:
24
G ( s) H ( s) =
s ( s + 1)( s + 6)
1. The steady state error is less than to /10 when the input is a ramp with a slope of 2 rad/sec
92
2 2
lim lim
10 24 4K
s ( s + 1)( s + 6)
As a result K 0.2
2. The phase margin is between 40 to 50 degrees
In this part of the solution, MATLAB sisotool can be very helpful. More detailed instructions
on using MATLAB sisotool is presented in the solutions for this particular problem. Similar
guidelinescouldbeusedforsimilarquestionsofthischapter.
SISOTOOLquickinstructions:
Once opening the sisotool by typing sisotool in MATLAB the command window (or in the m code), the
followingwindowpopsup:
WhereyoucaninserttransferfunctionsforC,G,H,andF,oryoucanleavesomeofthemasdefaultvalue(1),by
clickingonSystemData.Onceyousubstitutetransferfunctions,youwillseeagraphincludingtherootlocus
diagram, a closed loop Bode diagram, and an open loop Bode diagram indicating the Gain Margin and Phase
Marginaswell.
**Youcandragtheopenloopbodemagnitudediagramupanddowntoseetheeffectofgainchangeonallof
thegraphs.Sisotoolupdatesallthesegraphinstantly.Youcanalsodragthepolesandzerosontherootlocus
diagramtoobservetheeffectontheotherdiagrams.
93
YoucanalsoopenAnalysisPlotstabtoaddothergraphssuchasNyquistdiagramasshowninthefollowing
figure:
94
Inthisparticularquestion,youneedtoaddazerotoincludetheeffectof K d .Youcanaddazerobyusingthe
CompensatorEditortab,asshowninthefollowinggraph.Thelastthingyouneedtodoforthisproblem,isto
dragthelocationofzeroandgaininthefollowingdiagram(oredittheselocationsbyassigningCgainandZero
locationintheCompensatorEditortab),soitsatisfiesthePMof40to50deg;whilegainofCiskeptabove
0.2( K p 0.2 ,frompart(a)).
InthefollowingsnapshotofCompensatorEditor,Cgainor K p issetto1,andthezerolocationissetto2.7,
resultingin44.2[deg]phasemargin,presentedinthefollowingfigure.
95
60
2000
40
1000
20
Imag Axis
Magnitude (dB)
0
0
-1000
-20
-2000
-40 G.M.: Inf
-6 -5 -4 -3 -2 -1 0 Freq: Inf
Stable loop
Bode Editor for Closed Loop 1 (CL1)
-60
20
-90
Magnitude (dB)
-20
-40 -120
Phase (deg)
-60
0
Phase (deg)
-45 -150
-90
P.M.: 44.2 deg
-135
Freq: 2.07 rad/sec
-180 -180
-1 0 1 2 -2 -1 0 1 2
10 10 10 10 10 10 10 10 10
Finalanswers: K p = 1, K d = 0.3704
96
PreliminaryMATLABcodefor94:
%solving for k:
syms kc
omega=1.5
sol=eval(solve('0.25*kc^2=0.7079^2*((-0.25*omega^3+omega)^2+(-
0.375*omega^2+0.5*kc)^2)',kc))
%ploting bode with K=1.0370
s = tf('s')
K=1.0370;
num_G_a= 0.5*K;
den_G_a=s*(0.25*s^2+0.375*s+1);
CL_a = G_a/(1+G_a)
bode(CL_a);
sisotool;
Alternatively we can use ACSYS.
97
K p = 1, K d = 0.3704
98
99
9-5)(a)RamperrorConstant:
1000 ( K P + K D s ) 1000 K P
K v = lim s = = 100 K P = 1000 Thus K P = 10
s 0 s ( s + 10) 10
s + (10 + 1000 K D ) s + 1000 K P = 0

2
n = 1000 K P = 10000 = 100 rad/sec 2 n = 10 + 1000 K D = 2 0.5 100 = 100
90
Thus KD = = 0.09
1000
910
95(b)For K v = 1000 and = 0.707 ,andfrompart(a), n = 100 rad/sec,
131.4
2 n = 10 + 1000 K D = 2 0.707 100 = 141.4 Thus KD = = 0.1314
1000
(c)For K v = 1000 and = 1.0 ,andfrompart(a), n = 100 rad/sec,
190
2 n = 10 + 1000 K D = 2 1 100 = 200 Thus KD = = 0.19
1000
9-6)Theramperrorconstant:
1000 ( K P + K D s )
K v = lim s = 100 K P = 10, 000 Thus K P = 100
s 0 s ( s + 10)
1000 (100 + K D s )
Theforwardpathtransferfunctionis: G(s) =
s ( s + 10)
KD PM(deg) GM Mr BW(rad/sec) Maxovershoot(%)
0 1.814 13.5 493 46.6
0.2 36.58 1.817 525 41.1
0.4 62.52 1.291 615 22
0.6 75.9 1.226 753 13.3
0.8 81.92 1.092 916 8.8
1.0 84.88 1.06 1090 6.2
Thephasemarginincreasesandthemaximumovershootdecreasesmonotonicallyas K D increases.
911
Sample MATLAB CODE for time frequency responses

clear all
KD=1.0;
num = [-100/KD];
den = [0 -10];
G=zpk(num,den,1000);
figure(1)
margin(G)
M=feedback(G,1)
figure(2)
step(M);
figure(3)
bode(M)
Zero/pole/gain:
1000(s+100)
(s+111.3)(s+898.7)
Phase margin is Pm=180-95=85 Deg
Gm=
912
UsethecursortoobtainthePOandtrvalues.
KDincreaseresultsintheminimumovershoot.
Bandwidth is 1090 rads/s.
913
Use ACSYS to find the same results:
914
9-7)
PDcontrollerdesign
Theopenlooptransferfunctionofasystemis:
1
G(s) H (s) =
(2 s + 1)( s + 1)(0.5s + 1)
Thesolutionisverysimilarto94.Thetransferfunctionsareinsertedintosisotool,whereanotherrealzerois
addedtorepresenttheeffectof K d .Thatis C ( s ) = K p + K d s = K p (1 + K d s / K p ) ,whichiscalledthe
compensatortransferfunctioninsisotool.Theplaceofrealzerois Z = K p / K d ,andthegainofthe
compensatorisequalto K p ,asnotedinthefollowingsisotoolwindow:
915
Byfixingthegainto9,andstartingtochangethezerolocation,PMcanbeadjustedtoabove25[deg]as
requiredbythequestion.Thecurrentsettinghasazeroat3.5,whichresultedin30[deg]phasemarginand
33.1dBgainmarginasseeninthefollowingdiagrams.
Thedesignrequires K p = 9 and K d = K p / Z = 9 / 3.6 = 2.5
15 20
10
10
5
0
0
-10
-5
-20
-10
G.M.: 33.1 dB
-15 -30
Freq: 10.8 rad/sec
-4 -3 -2 -1 0 1
Stable loop
50
0
0
-45
-50
-90
-100
0 -135
-90
-180
-180 P.M.: 29.8 deg
Freq: 1.77 rad/sec
-270 -225
-1 0 1 2 -2 -1 0 1 2
10 10 10 10 10 10 10 10 10
s = tf('s')
Kp = 1
num_GH= Kp*1;
den_GH=(2*s+1)*(s+1)*(0.5*s+1);
916
GH=num_GH/den_GH;
CL = GH/(1+GH)
Sisotool
9-8: PD controller design: The open-loop transfer function of a system is:
60
G ( s) H ( s) =
s (0.4 s + 1)( s + 1)( s + 6)
(a) Design a PD controller to satisfy the following specifications:
(i) Kv =10
(ii) the phase margin is 45 degrees.
60
lim lim 10 10
0.4 1 1 6
As a result: 1
Therestoftheprocedureissimilarto97:
917
Thetransferfunctionsareinsertedintosisotool,whereanotherrealzeroisaddedtorepresenttheeffectof K d
.Thatis C ( s ) = K p + K d s = K p (1 + K d s / K p ) ,whichiscalledthecompensatortransferfunctioninsisotool.
Theplaceofrealzerois Z = K p / K d ,andthegainofthecompensatorisequalto K p ,asnotedinthe
sisotoolwindow:
K p = 1 fixedto1,andzerolocationwaschangedintheentirerealaxis.However,2oftheclosedlooppoles
remainedintherighthandsideofSplaneintherootlocusdiagram,indicatinginstabilityforallKdvalues.
Solutionfor K v = 10 withPDcontrollerandPM=45[deg]doesnotexist.Unstablecloselooppolesare
indicatedintherootlocusdiagramofthefollowingfigure:
918
98)
s = tf('s')
Kp = 1
num_GH= Kp*60;
den_GH=s*(0.4*s+1)*(s+1)*(s+6);
GH=num_GH/den_GH;
CL = GH/(1+GH)
sisotool
919
4500 K ( K D + K P s )
G ( s ) = Gc ( s )G p ( s ) =
s ( s + 361.2)
4500 KK P
RampErrorConstant: K v = lim sG ( s ) = = 12.458 KK P
s 0 361.2
1 0.0802
e ss = = 0.001Thus KK P 80.2 Let K P = 1 and K = 80.2
Kv KK P
AttributesofUnitstepResponse:
KD tr(sec) ts (sec) MaxOvershoot(%)
0 0.00221 0.0166 37.1
0.0005 0.00242 0.00812 21.5
0.0010 0.00245 0.00775 12.2
0.0015 0.0024 0.0065 6.4
0.0016 0.00239 0.00597 5.6
0.0017 0.00238 0.00287 4.8
0.0018 0.00236 0.0029 4.0
0.0020 0.00233 0.00283 2.8
Select K D 0.0017
(b)BWmustbelessthan850rad/sec.
920
KD GM PM(deg) Mr BW(rad/sec)
0.0005 48.45 1.276 827
0.0010 62.04 1.105 812
0.0015 73.5 1.033 827
0.0016 75.46 1.025 834
0.0017 77.33 1.018 842
0.00175 78.22 1.015 847
0.0018 79.07 1.012 852
Select K D 0.00175 .Alarger K D wouldyieldaBWlargerthan850rad/sec.
921
9-10)
TheforwardpathTransferFunction:N=20
200 ( K P + K D s )
G(s) =
s ( s + 1)( s + 10)
Tostabilizethesystem,wecanreducetheforwardpathgain.Sincethesystemistype1,reducingthe
gaindoesnotaffectthesteadystateliquidleveltoastepinput.Let K P = 0.05
200 ( 0.05 + K D s )
G(s) =
s ( s + 1)( s + 10)
UnitstepResponseAttributes:
KD ts(sec) MaxOvershoot(%)
0.01 5.159 12.7
0.02 4.57 7.1
0.03 2.35 3.2
0.04 2.526 0.8
0.05 2.721 0
0.06 3.039 0
0.10 4.317 0
When K D = 0.05 therisetimeis2.721sec,andthestepresponsehasnooversho
922
9-11)
(a)For e ss = 1,
200 ( K P + K D s )
K v = lim sG ( s ) = lim s = 20 K P = 1 Thus K P = 0.05
s 0 s 0 s ( s + 1)( s + 10)
200 ( 0.05 + K D s )
G ( s ) =
s ( s + 1)( s + 10)
AttributesofFrequencyResponse:
KD PM(deg) GM(deg) Mr BW(rad/sec)
0 47.4 20.83 1.24 1.32
0.01 56.11 1.09 1.24
0.02 64.25 1.02 1.18
0.05 84.32 1.00 1.12
0.09 93.80 1.00 1.42
0.10 93.49 1.00 1.59
0.11 92.71 1.00 1.80
0.20 81.49 1.00 4.66
0.30 71.42 1.00 7.79
0.50 58.55 1.03 12.36
Formaximumphasemargin,thevalueof K D is0.09.PM=93.80deg.GM= , M r = 1,
andBW=1.42rad/sec.
923
(b)SensitivityPlots:
M
ThePDcontrolreducesthepeakvalueofthesensitivityfunction SG ( j )
924
9-12)
PDcontrollerdesign:The open loop transfer function of a system is:
100
G ( s) H ( s) =
s (0.1s + 1)(0.02 s + 1)
Design the PD controller so that the phase margin is greater than 50 degrees and the BW is greater than
20 rad/sec.
The transfer functions are generated and imported in sisotool as in 9-4:
MATLABcode:
s = tf('s')
Kp = 1
num_GH= Kp*100;
den_GH=s*(0.1*s+1)*(0.02*s+1);
GH=num_GH/den_GH;
CL = GH/(1+GH)
sisotool
Followingsimilarstepsin94,thelooptransferfunctionsareinsertedintosisotool.Anotherrealzeroisaddedto
representtheeffectof K d .Thatis C ( s ) = K p + K d s = K p (1 + K d s / K p ) ,whichiscalledthecompensator
transferfunctioninsisotool.Theplaceofrealzerois Z = K p / K d ,andthegainofthecompensatorisequal
to K p .Thezerolocationand K p gainwerechangedinteractivelyinsisotooluntilthedesiredPM(59.5[deg])
andBWisachieved.FollowingfiguresshowsthisPMatcrossoverfrequencyof38.6rad/sec,whichinsuresBW
925
of higher than 38.6 rad/sec (as the bandwidth is @ 3dB rather than 0 DB, i.e. bandwidth occurs at higher
frequencycomparedtocrossoverfrequency).
Finalpossibleanswer: K p = 0.25 and K d = K p / Z = 0.25 / 5 = 0.05
926
927
9-13)
Leadcompensatordesign:
1000 K
G(s) H (s) =
s (0.2 s + 1)(0.005s + 1)
Designacompensatorsuchthatthesteadystateerrortotheunitstepinputislessthan0.01andtheclosedloop
dampingratio>0.4.
Thetransferfunctionsaregeneratedandimportedinsisotoolasin94:
1
e ss = < 0.01 1000 K p > 101 Therefore, K p isselectedas150:( K p =0.11)
1 + 1000 K p
Toachievetherequireddampingratio,thepolesoftheclosedloopsystemareplacedwithanangleoflessthan
ArcCos(=0.4),intherootlocusdiagramofsisotool.Thisisdonebyiterativelychangethelocationofpolesand
zeros of a lead compensator and setting K p =0.11. The pole and zero (which perform as a lead compensator
when the pole is further away from zero to the left) are inserted in sisotool as explained in 94. The lead
compensatorwillintroducesomephaseleadatlowerfrequenciesaboutthezerolocationwhichimprovesthe
closed loop response in terms of damping and phase margin. Following is the chosen location for lead
compensatorpoleandzero:
Pole@100rad/sec
Zero@1.08rad/sec
Whichresultedinsmallestanglesof dominant polelocations (theonesclosertoimaginaryaxis)with thereal
axis.Thissmallanglemeanshigherdampingofthepolesas=ArcCos(polesanglewithrealaxis).
928
50
2
0
1
-50
0
-100
-1
-150
-2
G.M.: 17.3 dB
-200
Freq: 3.64 rad/sec
-2 -1 0 1
Stable loop
-90
0
-135
-100
-180
-200
-225
0
-90 -270
-180
-315 P.M.: 56.7 deg
-270
Freq: 0.975 rad/sec
-360 -360
-2 0 2 4 -2 0 2 4
10 10 10 10 10 10 10 10

929
Inthisparticularcase,closedloopcomplexpolescanbeobservedintheshownrootlocusdiagramatabout
0.81.2j.Thiscorrespondstodampingofabout:
cos(atan(1.2/0.8))=0.554 0.55
s = tf('s')
Kp = 1
num_GH= Kp*60;
den_GH=s*(0.4*s+1)*(s+1)*(s+6);
GH=num_GH/den_GH;
CL = GH/(1+GH)
figure(1)
margin(CL)
sisotool
930
9-14)
PDcontrollerdesignedforamaximumovershootandamaximumsteadystateerror
1
0.005
lim
Therefore:
1 1 4
0.005 K p > Kp >
250 K p 0.005 * 250 5
Let K p = 1 ,then:

M p = exp( ) < 0.20 > 0.45
2
1

Let 0.6;then:
250 1 1250 1

0.2 250 1 250 1250 5 1250
Accordingly, 1250,or 35.35.therefore,

1250 5 2 1250 2 0.6 0.35
whichgives: 0.034

Thedesigncharacteristicscanbeobservedinthediagrambelow:
931
1 110
100
0.5
90
0
80
-0.5
70
G.M.: Inf
-1 60
Freq: NaN
-4 -3 -2 -1 0
Stable loop
Bode Editor for Closed Loop 1 (CL1) x 104 50
0
0
P.M.: 90 deg
-10 Freq: 3.68e+004 rad/sec
-20 -45
-30
0
-90
-45
-90 -135
-2 0 2 4 6 -4 -2 0 2
10 10 10 10 10 10 10 10 10
9-15)
Leadcompensatorcontrollerdesign
Thebodediagramofthesystemwithoutleadcompensatorisshownbelow:
932
Bode Diagram
20
Magnitude (dB) 0
-20
-40
-60
-80
0
-45
Phase (deg)
-90
-135
-180
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)

IndicatingaPMof32.7[deg].Toreach45[deg]phasemargin,additional12.3[deg]phaseleadis
needed.At=17.3rad/scrossoverfrequency,ifin=12.3then
1 sin
r 0.6488
1 sin
As= 17.3,then=0.0718

Sincethegainisloweredby 1.2838
.
Againcompensatorwithgainof0.7789isrequired,where,
1

1
933
9-16) If r = 0.1 sin 55
tan tan 17.3 tan 17.3 tan 1.73 12.3
then from trial and error we found = 0.014088 and required gain would be 9.7185
9-17) K lim sG s G s 100K 100 K 1
First plot the bode diagram of uncompensated system when K = 1
Bode diagram for Loop T.F. is included for K=1
MATLAB CODE:
s = tf('s')
Kp = 1
num_GH= Kp*100;
den_GH=s*(0.2*s+1)*(0.05*s+1);
GH=num_GH/den_GH
lag_tf=(s/2+1)/(s/0.2+1)
lead_tf=(s/4+1)/(s/50+1)
LL=lag_tf*lead_tf
OL=GH*LL
CL =OL/(1+OL);
figure(1)
margin(GH)
figure(2)
margin(OL)
figure(3)
bode(CL)
grid on;
934
Bode Diagram
Gm = -12 dB (at 10 rad/sec) , Pm = -28.3 deg (at 18.8 rad/sec)
100
50
Magnitude (dB)
-50
-100
-150
-90
-135
Phase (deg)
-180
-225
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
The bode diagram with K=1 shows -28 deg PM at 18.8 rad/sec.
According to the requirements the gain must be greater than or 250 for 0.2 / and
.
must be less than or 0.01 for 200 /
In order to achieve above requirements, a lead-lag compensator will be appropriate.
Using a lag compensator will allow lower gain at frequencies less then 1 and using a lead
compensator will allow to increase phase margin
For the lag compensator, 1/10 is chosen to boost the low frequency amplitude
1 /2 1
1 /0.2 1
In order to introduce some phase lead to obtain the require PM, a lead compensator is also
designed as:
1 1
0.0112
10 28.310
935
Where is overlaid with the crossover frequency (28.3 rad/sec) for applying the maximum
phase lead at this frequency. The Lead compensator T.F. will be as follows:
1 /5 1
1 /50 1
Resulting in the following Bode diagram for the compensated system, showing 44 deg PM:
To obtain a slightly higher PM, lead compensator zero was re-tuned, where the zero is pulled
closer to imaginary axis from -5 to -4:
1 /4 1
1 /50 1
This resulted in a higher PM as shown in the following bode diagram of loop transfer function:
Bode Diagram
Gm = 14.5 dB (at 30.7 rad/sec) , Pm = 45 deg (at 10.6 rad/sec)
100
50
Magnitude (dB)
-50
-100
-90
-135
Phase (deg)
-180
-225
-270
-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)
Correspondingly, the Bode diagram of closed loop system can be shown as:
936
Bode Diagram
20
Magnitude (dB) -20
-40
-60
-80
-100
0
-90
Phase (deg)
-180
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
9-18)SeeChapter5solutionsforMATLABcodesforthisproblem.
100 K P +
KI
100 ( K P s + K I )
G ( s ) =
s
For K v = 10, K v = lim sG ( s ) = lim s = K I = 10
s + 10 s + 100
2
s 0 s 0
(
s s + 10 s + 100
2
)

Thus K I = 10.
(b)Letthecomplexrootsofthecharacteristicequationbewrittenas s = + j15 and s = j15.
Thequadraticportionofthecharacteristicequationis
2
(
s + 2 s + + 225 = 0
2
)
937
s + 10 s + (100 + 100 K P ) s + 1000 = 0

3 2
Thecharacteristicequationofthesystemis
Thequadraticequationmustsatisfythecharacteristicequation.Usinglongdivisionandsolveforzero
remaindercondition.
s + (10 2 )
s + 2 s + + 225 s + 10 s + (100 + 100 K P ) s + 1000

2 2 3 2
s + 2 s + + 225 s
3 2
( 2
)
(
(10 2 ) s + 100 K P 125 s + 1000
2
) 2

(10 2 ) s + ( 20 4 ) s + (10 2 ) ( s + 225 )
2 2 2
(100 K P )
+ 3 20 125 s + 2 10 + 450 1250
2 3 2
3 2
Forzeroremainder, 2 10 + 450 1250 = 0 (1)
2
and 100 K P + 3 20 125 = 0 (2)
TherealsolutionofEq.(1)is = 2.8555 .FromEq.(2),
2
125 + 20 3
KP = = 15765
.
100
Thecharacteristicequationrootsare: s = 2.8555 + j15, 2.8555 j15, and s = 10 + 2 = 4.289
(c)RootContours:
938
100 K P s 100 K P s
Geq ( s ) = =
s + 10 s + 100 s + 1000
3 2
( s + 10 ) ( s 2
+ 100 )
RootContours:
939
9-19)

100 K P +
KI
100 ( K P s + K I )
G ( s ) =
s For K = 10, K v = lim sG ( s ) = lim s = K I = 10
s + 10 s + 100
2 v
s 0 s 0
(
s s + 10 s + 100
2
)
Thustheforwardpathtransferfunctionbecomes
10 (1 + 0.1K P s )
G(s) =
(
s 1 + 0.1s + 0.01s
2
)
AttributesoftheFrequencyResponse:
KP PM(deg) GM(dB) Mr BW(rad/sec)
0.1 5.51 1.21 10.05 14.19
0.5 22.59 6.38 2.24 15.81
0.6 25.44 8.25 1.94 16.11
0.7 27.70 10.77 1.74 16.38
0.8 29.40 14.15 1.88 16.62
940
0.9 30.56 20.10 1.97 17.33
1.0 31.25 2.00 18.01
1.5 31.19 1.81 20.43
1.1 31.51 2.00 18.59
1.2 31.40 1.97 19.08
When K P = 11
. and K I = 10, K v = 10, thephasemarginis31.51deg.,andismaximum.
Thecorrespondingrootsofthecharacteristicequationrootsare:
5.4, 2.3 + j13.41, and 2.3 j13.41
ReferringtheserootstotherootcontoursinProblem108(c),thecomplexrootscorrespondsto
arelativedampingratiothatisnearoptimal.
(b)SensitivityFunction:
941
Inthepresentcase,thesystemwiththePIcontrollerhasahighermaximumvalueforthesensitivity
function.
942
9-20)
100 ( K P s + K I )
G(s) =
(
s s + 10 s + 100
2
)
For K v = 100,
100 ( K P s + K I )
K v = lim sG ( s ) = lim s = K I = 100 Thus K I = 100.
s 0 s 0
(
s s + 10 s + 100
2
)
s + 10 s + (100 + 100 K P ) s + 100 K I = 0
3 2
(b)Thecharacteristicequationis
RouthTabulation:
3
s 1 100 + 100 K P
2
s 10 10,000
Forstability,100 K P 900 > 0 Thus KP > 9
1
s 100 K P 900 0
0
s 10,000
RootContours:
100 K P s 100 K P s
Geq ( s ) = 3 2
=
s + 10 s + 100 s + 10,000 ( s + 23.65)( s 6.825 + j19.4)( s 6.825 j19.4)
943
(c) K I = 100
100 ( K P s + 100 )
G(s) =
(
s s + 10 s + 100
2
)
Thefollowingmaximumovershootsofthesystemarecomputedforvariousvaluesof K P .
KP 15 20 22 24 25 26 30 40 100 1000
ymax 1.794 1.779 1.7788 1.7785 1.7756 1.779 1.782 1.795 1.844 1.859
WhenKP=25,minimumymax=1.7756
944
9-21)
100 ( K P s + K I ) 100 K I
G(s) = For Kv = = 10, K I = 10
(
s s + 10 s + 100
2
) 100
(b)CharacteristicEquation: (
s3 + 10 s 2 + 100 K + 1 s + 1000 = 0
P )
RouthTabulation:
3
s 1 100 + 100 K P
2
s 10 1000
Forstability,KP>0
1
s 100 K P 0
0
s 1000
RootContours:
100 K P s
Geq ( s ) = 3 2

s + 10 s + 100 s + 1000
945
(c)Themaximumovershootsofthesystemfordifferentvaluesof K P rangingfrom0.5to20are
computedandtabulatedbelow.
KP 0.5 1.0 1.6 1.7 1.8 1.9 2.0 3.0 5.0 10 20
ymax 1.393 1.275 1.2317 1.2416 1.2424 1.2441 1.246 1.28 1.372 1.514 1.642
WhenKP=1.7,maximumymax=1.2416
K
9-22) K lim sG s G s 20 20 K 5
let K = 6 and targeted PM = 45o . To include some integral action, Ki is set to 1.
First , lets take a look at uncompensated system:
The open loop bode shows as PM of -25.2 @ 4.35 rad/sec:
946
Bode Diagram
Gm = -12.6 dB (at 2.2 rad/sec) , Pm = -25.2 deg (at 4.35 rad/sec)
100
50
Magnitude (dB)
-50
-100
-150
-135
-180
Phase (deg)
-225
-270
-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)
To achieve the PM of 45 deg, we need to add a phase lead of (45 (-25.2)=70.2). By try and error, 2
compensators (a double lead compensator) each with phase lead of 55 deg was found suitable.
Considering the change in cross over frequency after applying the lead filters, overall, a PM of 52 deg
was obtained as seen in the bode diagram of compensated loop:
Double Lead filter design:
1 sin 1 sin 55 1 1
10.0590, 0.0210
1 sin 1 sin 55 10.059 1510.0590
The maximum phase lead of the compensators are placed at 15 rad/sec, which resulted in a larger PM
(=52.5 deg) compared to applying this phase lead at original cross over frequency of 4.35 rad/sec. This
was due to the shape of phase diagram affected by integral action (i.e. phase starts at -180 @ = 0
rad/sec.
The gain crossover frequency is 6.92 rad/sec. Bode diagram of compensated loop transfer
function can be observed in the following figure, showing a PM pf 52.5 deg:
947
Bode Diagram
100
Magnitude (dB) 50
-50
-100
-90
-135
Phase (deg)
-180
-225
-270
-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)
Bode Diagram
20
0
Magnitude (dB)
-20
-40
-60
-80
-100
0
-90
Phase (deg)
-180
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
948
MATLAB code:
s = tf('s')
Kp = 6;
Ki = 1;
num_GH= 24*(Kp+Ki/s);
den_GH=s*(s+1)*(s+6);
GH=num_GH/den_GH;
%lead design
PL=55
CRover=15
alpha=(1+sin(PL/180*pi))/(1-sin(PL/180*pi))
T=1/alpha^0.5/CRover
lead=(1+T*alpha*s)/(1+T*s)
LT=GH*lead*lead %double lead compensation
CL = LT/(1+LT);
figure(1)
Margin(GH)
figure(2)
Margin(LT)
figure(3)
Bode(CL)
grid on;
9-23)
1
e 0.05
K
40 K s K I
K lim s G s G s lim KI 20
s 2 s 20
Lets consider 21
As gain crossover frequency is 1 |G G j | 1 1.25
Lets see if the PM is in the required range. The bode of the loop transfer function shows a PM
of -57 deg at 3.31.
949
Bode Diagram
Gm = -Inf dB (at 0 rad/sec) , Pm = -57.1 deg (at 3.31 rad/sec)
100
50
Magnitude (dB)
-50
-100
-180
Phase (deg)
-225
-270
-1 0 1 2
10 10 10 10
Frequency (rad/sec)
By try and error, a double lead compensator, each with phase lead of 53 deg was found suitable.
Considering the change in cross over frequency after applying the lead filters, overall, a PM of 35.4 deg
1 sin 1 sin 53 1 1
8.9322, 0.0446
1 sin 1 sin 53 8.9322 7.58.9322
The maximum phase lead of the compensators are placed at 7.5 rad/sec, resulting in a larger PM (= 52.5
deg) compared to applying this phase lead at original cross over frequency of 3.31 rad/sec. This was due
to the shape of phase diagram affected by integral action (i.e. phase starts at -180 @ = 0 rad/sec.
Then the gain crossover frequency is 6.8 rad/sec. Bode diagram of compensated loop can be
observed in the following figure, showing a PM pf 35.4 deg:
950
MATLAB code: Bode Diagram

s = tf('s') 100
Kp = 1.25;
Ki = 21; 50
Magnitude (dB)
den_GH=s*(s+2)*(s+20); 0
GH=num_GH/den_GH;
CL = GH/(1+GH);
-50
%lead design
PL=53
-100
CRover=7.5
alpha=(1+sin(PL/180*pi))/(1
-150
-sin(PL/180*pi)) -135
LT=GH*lead*lead -180
Phase (deg)
CL = LT/(1+LT);
figure(1) -225
Margin(GH)
figure(2)
Margin(LT) -270
-1 0 1 2 3
figure(3) 10 10 10 10 10
Bode(CL) Frequency (rad/sec)
grid on;
Bode Diagram
50
0
Magnitude (dB)
-50
-100
-150
0
-90
Phase (deg)
-180
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
951
9-24)
To satisfy the unity DC gain (G(0)/(1+Kp*G(0)) = 1), Kp should be equal to 1: Kp=1
In order to add some integral action, 0.2 was chosen as the integral gain.
First, the bode plot of the Loop transfer function is obtained demonstrating a PM of -23.8 deg at 3
rad/sec cross over frequency:
Bode Diagram
Gm = -9.18 dB (at 1.82 rad/sec) , Pm = -23.8 deg (at 3 rad/sec)
100
50
Magnitude (dB)
-50
-100
-135
-180
Phase (deg)
-225
-270
-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/sec)
First, the bode plot of the Loop transfer function is obtained demonstrating a PM of -23.8 deg at 3
rad/sec cross over frequency:
By try and error, a double lead compensator, each with phase lead of 48 deg was found suitable.
Considering the change in cross over frequency after applying the lead filters, overall, a PM of 41.9 deg
1 sin 1 sin 48 1 1
6.7865, 0.0427
1 sin 1 sin 48 6.7865 98.9322
952
The maximum phase lead of the compensators are placed at 9 rad/sec, resulting in a larger PM (= 41.9
deg) compared to applying this phase lead at original cross over frequency of 3 rad/sec. This was due to
the shape of phase diagram affected by integral action (i.e. phase starts at -180 @ = 0 rad/sec.
Then the gain crossover frequency is 4.35 rad/sec. Bode diagram of compensated loop can be
observed in the following figure, showing a PM pf 35.4 deg:
Bode Diagram
100
50
Magnitude (dB)
-50
-100
-150
-90
-135
Phase (deg)
-180
-225
-270
-2 -1 0 1 2 3
10 10 10 10 10 10
Frequency (rad/sec)
MATLAB code:
s = tf('s')
Kp=1;
Ki=0.2;
den_GH=s*(5*s+7)*(s+3);
GH=num_GH/den_GH;
%lead design
PL=48
CRover=9
alpha=(1+sin(PL/180*pi))/(1-sin(PL/180*pi))
LT=GH*lead*lead
CL = LT/(1+LT);
953
figure(1)
Margin(GH)
figure(2)
Margin(LT)
figure(3)
Margin(CL)
Bode Diagram
50
0
Magnitude (dB)
System: CL
-50
-100
-150
0
-90
Phase (deg)
-180
-270
-1 0 1 2 3
10 10 10 10 10
Frequency (rad/sec)
The Bandwidth can be obtained from -3dB in magnitude diagram of the Bode plot. He above data point
in the figure shows BW = 7.44 rad/sec
954
9-25) 2 , therefore, 2
From bode plot of uncompensated loop, we have PM = -111 at = 0.363 rad/s:
Bode Diagram
Gm = -25.9 dB (at 0.0576 rad/sec) , Pm = -111 deg (at 0.363 rad/sec)
100
50
Magnitude (dB)
-50
-100
270
180
Phase (deg)
90
0
-3 -2 -1 0 1
10 10 10 10 10
Frequency (rad/sec)
K
A PI controller can be expressed as PI K sK . The effect is similar to adding a Zero at . Lets
K
place this zero at 71/500 to cancel the Phase lag originating from the unstable zero of G at +71/500:
2353 K (71 500 s )
( G (s) = ).
71s ( 40 s + 13)(5000 s + 181)
The compensator can be expressed as:

K
PI K s K s 71/500 , where K can be adjusted in sisotool as the overall gain of the
K
loop, until the required PM is achieved. At K 37, PM=50 deg as seen in the following sisotool
results:
955
10 50
0
5
-50
0 -100
-150
-5
-200
G.M.: 10.5 dB
-10 -250 Freq: 0.0343 rad/sec
-10 -5 0 5 10
Stable loop
200
270
0
180
-200
90
-400
360 0
0 -90
P.M.: 50.5 deg
Freq: 0.0137 rad/sec
-360 -180
-4 -2 0 2 4 -4 -2 0 2 4
10 10 10 10 10 10 10 10 10 10
Considering the slow dynamic of the ship, and the RHS zero, the crossover frequency is relatively low.
The root locus diagram can be seen as:
Zoom in
956
Root Locus
0.06
0.04
0.02
Imaginary Axis
-0.02
-0.04
-0.06
-0.08
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0.1 0.2
Real Axis
Zoom out
Root Locus
0.6
0.4
0.2
Imaginary Axis
-0.2
-0.4
-0.6
-0.8
-8 -7 -6 -5 -4 -3 -2 -1 0
Real Axis
MATAB Code:
s = tf('s')
Kp=1;
Ki=1;
num_G= 2353*2*(71-500*s)
den_G=71*s*(40*s+13)*(5000*s+181)*(71+500*s)^2;
G=num_G/den_G;
%PI design
957
Kp=1
Ki=71/500
PI=Kp+Ki*s
figure(100)
Margin(G)
figure(101)
rlocus(G*PI)
sisotool
9-26) a) Transfer functions G and H are generated in MATLAB and imported into sisotool:
MATLAB Code:
s = tf('s')
num_G= 2*10^5;
den_G=s*(s+20)*(s^2+50*s+10000);
G=num_G/den_G;
num_c1= 0.05*s+1;
den_c1=0.01*s+1;
c1=num_c1/den_c1;
num_c2= s/0.316+1;
den_c2=s/3.16+1;
c2=num_c2/den_c2;
sisotool
(a) The gain was changed until the cross over frequency matches 31.6 rad/sec as a requirement. At
K=30.3, the desired cross over frequency of 31.6 rad/sec happens as can be seen in the following
sisotool results:
958
300 50
200
0
100
0
-50
-100
-200
-100
G.M.: 5.26 dB
-300 -200 -100 0 100 200
Stable loop
50
-90
-50 -180
-100
0
-270
-180
P.M.: 62.5 deg
Freq: 31.6 rad/sec
-360 -360
0 1 2 3 0 1 2 3 4
10 10 10 10 10 10 10 10 10
959
. K
b) K lim sG s G s K Lim s K 30.3 , then
.
K 30.3
c) Again, K lim sG s G s K Lim s r. To have

K 100 , the overall gain of the PI controller should be equal to 100 (r = 100).
d & e) In this part, the PI pole is asked to be placed at -3.16 rad/sec and the crossover frequency needs to
be at 31.6 rad/sec. The zero and the gain of the PI controller needs to be designed.
r (s + 1)
Considering the structure of the PI controller given in the question, H ( s ) = , the corresponding
(rs + 1)
pole is set to -3.16 in sisotool. The place of the zero and the overall gain is iteratively changed in the
MATLAB sisotool to achieve the crossover frequency of 31.6 rad/sec.
With a zero at -0.06321 rad/sec and overall gain of K=270, required crossover frequency (31.6 rad/sec)
and PM of 21.9 deg is obtained as shown in the following sisotool results:
300 150
200 100
100 50
0 0
-100 -50
-200 -100
G.M.: 6.61 dB
-300 -150 Freq: 53.2 rad/sec
-300 -200 -100 0 100 200
Stable loop
100
-90
-100 -180
-200
0
-270
-180
P.M.: 21.9 deg
Freq: 31.6 rad/sec
-360 -360
-2 0 2 4 -4 -2 0 2 4
10 10 10 10 10 10 10 10 10
960
e) the presented sisotool figure shows the compensated bode diagram and 21.9 deg of PM.
961
9-27)
KDs + KPs + KI KI 2
2
KI
Gc ( s ) = K P + K D s + = = (1 + K D 1 s ) K P 2 +
s s s
where
K P = K P 2 + K D1 K I 2 K D = K D1 K P 2 KI = KI 2
100 (1 + K D1 s )( K P 2 s + K I 2 )
G ( s ) = Gc ( s )G p ( s ) = K v = lim sG ( s ) = K I 2 = 100
(
s s + 10 s + 100
2
) s 0
Thus
K I = K I 2 = 100
ConsideronlythePIcontroller,(with K D1 = 0 )
ForwardpathTransferFunction: CharacteristicEquation:
100 ( K P 2 s + 100 )
G ( s ) = s + 10 s + (100 + 100 K P 2 ) s + 10, 000 = 0
3 2

(
s s + 10 s + 100
2
)

Forstability, K P 2 > 9 .Select K P 2 = 10 forfastrisetime.
1000 (1 + K D1 s )( s + 10 )
G(s) =
(
s s + 10 s + 100
2
)
962
When K D1 = 0.2, therisetimeand

overshootrequirementsaresatisfied.
K D = K D1 K P 2 = 0.2 10 = 2
K P = K P 2 + K D1 K I 2 = 10 + 0.2 100 = 30
100
Gc ( s ) = 30 + 2 s +
s
UnitstepResponse
963
9-28)
ProcessTransferFunction:
0.2 s
Y (s) e 1
Gp (s) = =
U (s) 1 + 0.25s (1 + 0.25s ) (1 + 0.2 s + 0.02 s 2
)
(a)PIController:
KI
KP +
s 200 ( K P s + K I )
G ( s ) = Gc ( s )G p ( s ) =
(1 + 0.25s ) ( 1 + 0.2 s + 0.02 s
2
)
s ( s + 4 ) s + 10 s + 50
2
( )
200 K I
For K v = 2, K v = lim sG ( s ) = = KI = 2 Thus K I = 2
s 0 4 50
200 ( 2 + K P s )
Thus G(s) =
(
s ( s + 4 ) s + 10 s + 50
2
)
Thefollowingvaluesoftheattributesoftheunitstepresponsearecomputedforthesystemwithvarious
valuesfor K P .
Maxovershoot ts ts
KP
(%) (sec) (sec)

0.1 19.5 0.61 2.08
0.2 13.8 0.617 1.966
0.3 8.8 0.615 1.733
0.4 4.6 0.606 0.898
0.5 1.0 0.5905 0.878
0.6 0 0.568 0.851
0.7 0 0.541 1.464
0.8 0 0.5078 1.603
1.0 0 0.44 1.618
964
965
966
967
9-29)
968
9-30) If there is no disturbance then
0.9 2
2
Lets consider PID controller as:
1 1 1/
If D is sufficiently smaller than I then the I has minor effect in PID controller. Lets examine
PID controller when D is varied.
0.5 180
100 90
If 10 65 0.5
20 1
so lets consider D 10, then 0.1

D
As I to be large enough with respect to D, let I 20 D 200 0.005

I
Now we have to determine the value of K so that the gain at the crossover frequency remains at 1.
If K=1 then |G s G s | . 20. Therefore, 20, or K=0.05
9-31) Let
1 1
1 1
where 0.01. therefore, 10 1 100
which gives 10 . Now 10,then PM 40 . As a result, 91o phase lead is

required to achieve PM 45
The crossover frequency is 12.3 rad/sec as can be seen in the uncompensated bode diagram. The
lag compensator must position 5rad/sec, where its gain is 17.5 dB. Therefore the ratio of
lag compensator can be chosen for this purpose as 3 10
969
Bode Diagram
40
20
Magnitude (dB)
-20
-40
0
-45
Phase (deg)
-90
-135
-180
-2 -1 0 1 2
10 10 10 10 10
Frequency (rad/sec)
Now the gain, which is obtained from combination of lead and lag compensator, is
1 5 1 5

1 5 1 5
or
1 5 1 5
20 20
1 5 1 5
where 1 for 1 < 1. Since it is required that the final gain is increased by 17.5 dB,
lets choose .
On the other hand, the corner frequency is 0.5 2. Therefore,
1 5
20 23.5
1 5
970
1 5
20 23.5 17.5 6
1 5
As a result, the actual phase reduction is 5.33
The required phase lead is 45 5.33 3 53.33, where

53.33
By trial and error, we can find 0.068 and 0.35. Therefore 147 where
971
9-32)
972
UnitstepResponse.
Theunitstepresponseshowsamaximumovershootof26%.Althoughtherelativedampingratioofthe
complexrootsis0.707,therealpoleofthethirdordersystemtransferfunctionisat0.667whichadds
totheovershoot.
(c)
0.00667 (1 + K D1 s )( K P 2 s + K I 2 )
G ( s ) = Gc ( s )G p ( s ) =
(
s s + 0.00667
2
)
For K v = 100, K I 2 = K I = 100. Letusselect K P 2 = 50. Then
0.00667 (1 + K D1 s )( 50 s + 100 )
G(s) =
(
s s + 0.00667
2
)
Forasmallovershoot, K D1mustberelativelylarge.When K D1 = 100, themaximumovershootis
approximately4.5%.Thus,
973
K P = K P 2 + K D1 K I 2 = 50 + 100 100 = 10050

K D = K D1 K P 2 = 100 50 = 5000
K I = 100
3 2
SystemCharacteristicEquation: s + 33.35 s + 67.04 s + 0.667 = 0
Roots: 0.01,2.138,31.2
UnitstepResponse.
974
9-33)(a)
Z ( s) 1 1 0.00667
G p ( s) = = 2
= 2
= 2

F ( s) Ms + K s 150 s + 1 s + 0.00667
Thetransferfunction G p ( s ) haspolesonthe j axis.Thenaturalundampedfrequencyis
n = 0.0816 rad/sec.
(b)PIDController:
G ( s ) = Gc ( s )G p ( s ) =
(
0.00667 K D s + K P s + K I
2
)
(
s s + 0.00667
2
)
K v = lim sG ( s ) = K I = 100 Thus K I = 100
s 0
s + 0.00667 K D s + 0.00667 (1 + K P ) s + 0.00667 K I = 0

3 2
2
For = 1 and n = 1rad/sec,thesecondordertermofthecharacteristicequationis s + 2 s + 1.
Dividingthecharacteristicequationbytheseondorderterm.
s + ( 0.00667 K D 2 )
s + 2 s + 1 s + 0.00667 K D s + ( 0.00667 + 0.00667 K P ) s + 0.00667 K I

2 3 2
s + 2s + s
3 2
( 0.00667 K D
2 ) s + ( 0.00667 K P 0.99333 ) s + 0.00667 K I
2
( 0.00667 K 2 ) s + ( 0.01334 K D 4 ) s + 0.00667 K D 2

2
D

( 0.00667 K P
0.01334 K D + 3.00667 ) s + 0.00667 K I 0.00667 K D + 2
Forzeroremainder,
0.00667 K P 0.01334 K D + 3.00667 = 0 (1)
0.00667 K D + 0.00667 K I + 2 = 0 (2)
FromEq.(2),
0.00667 K D = 0.00667 K I + 2 = 2.667 Thus K D = 399.85
FromEq.(1),
975
0.00667 K P = 0.01334 K D 3.00667 = 2.3273 Thus K P = 348.93
G(s) =
(
0.00667 399.85s + 348.93s + 100
2
)
(
s s + 0.00667
2
)
s + 2.667 s + 2.334 s + 0.667 = ( s + 1) ( s + 0.667 ) = 0

3 2 2

Roots:1,1,0.667
UnitstepResponse.
Themaximumovershootis20%.
976
9-34) a) As , therefore, or
b)
According to the second order system:
1.8 1.8
0.36
5
0.6
1
Let's first add a PD controller with G s 1 K D s, and find KD which satisfy the maximum
overshoot requirement.
After writing the closed-loop transfer function including the PD controller, the characteristic equation
(denominator of the closed loop T.F.) is:
Therefore:
1000 0.36 129.6
Also, 2 2 2 1000 0.6 0.36 50 382 Nsec/m
KI
Now lets add a PI controller with G s K , and find Kp and KI by using following table.
KI
Now, for the PI part, K I should be selected so that the additional pole at does not interfere
K
with the system dynamics. This pole is usually placed at least 1 decade lower (frequency wise)
than the slowest existing poles of the system. In this case, since 0.05, lets have
KI
0.005, resulting in K I K 0.005 129.6 0.005 0.648
K
The step response is obtained through the following MATLAB code, showing the rise time of
less than 5 sec, and almost no overshoot
977
Step Response
1
0.99
0.98
0.97
0.96
Amplitude
0.95
0.94
0.93
0.92
0.91
0.9
0 5 10 15 20 25 30 35 40 45 50
Time (sec)
MATLAB code:
s = tf('s')
Kp = 129.6
Kd = 382
Ki = 0.648
num_GH= (Kp*Kd*s)*(1+Ki/s);
den_GH=(1000*s+50);
GH=num_GH/den_GH;
CL = GH/(1+GH)
figure(1)
step(CL)
xlim([-1 50])
ylim([0.9 1])
978
9-35)(a)ProcessTransferFunction: ForwardpathTransferFunction
4 4 ( KP + KD s)
G p ( s) = 2
G ( s ) = Gc ( s )G p ( s ) = 2

s s
2 2
CharacteristicEquation: s + 4 K D s + 4 K P = s + 1.414 s + 1 = 0 for = 0.707, n = 1rad/sec
K P = 0.25 and K D = 0.3535
UnitstepResponse.
Maximumovershoot=20.8%
(b)Selectarelativelylargevaluefor K D andasmallvaluefor K P sothattheclosedlooppolesarereal.
Theclosedloopzeroat s = K P / K D isveryclosetooneoftheclosedlooppoles,andthesystem
dynamicsaregovernedbytheotherclosedlooppoles.Let K D = 10 andusesmallvaluesof K P .
Thefollowingresultsshowthatthevalueof K P isnotcriticalaslongasitissmall.
979
Kp Maxovershoot tr ts
(%) (sec) (sec)
0.1 0 0.0549 0.0707
0.05 0 0.0549 0.0707
0.2 0 0.0549 0.0707
(c)ForBW 40rad/secand M r = 1,wecanagainselect K D = 10 andasmallvaluefor K P .
Thefollowingfrequencydomainresultssubstantiatethedesign.
Kp PM Mr BW
(deg) (rad/sec)
0.1 89.99 1 40
0.05 89.99 1 40
0.2 89.99 1 40
980
9-36)(a)ForwardpathTransferFunction: CharacteristicEquation:
10, 000 ( K P + K D s ) 3 2
G ( s ) = Gc ( s )G p ( s ) = s + 10 s + 10,000 K D s + 10,000 K P = 0
s
2
( s + 10 )
RouthTabulation:
3
s 1 10,000 K D
2
s 10 10,000 K P
Thesystemisstablefor K P > 0 and K D > 0.1K P
1
s 10,000 K D 1000 K P 0
0
s 10,000 K P
(b)RootLocusDiagram:
10, 000 K P
G ( s ) =
s
2
( s + 10 )
981
RootContours:
0 K D < , K P = 0.001, 0.002, 0.005, 0.01.
10,000 K D s
Geq ( s ) = 3 2

s + 10 s + 10,000 K P
(c)Therootcontoursshowthatforsmallvaluesof K P thedesignisinsensitivetothevariationof K P .This

meansthatifwechoose K P tobebetween0.001and0.005,thevalueof K D canbechosentobe0.005
forarelativedampingratioof0.707.Let K P = 0.001 and K D = 0.005. Gc ( s ) = 0.001 + 0.005 s.
Theforwardpathtransferfunctionbecomes
982
10 (1 + 5 s )
G(s) =
s
2
( s + 10 )
Sincethezeroof G ( s ) isats=0.2,whichisveryclosetos=0,G(s)canbeapproximatedas:
50
G(s)
s ( s + 10 )
Forthesecondordersystem, = 0.707.UsingEq.(7104),therisetimeisobtainedas
2
1 0.4167 + 2.917
tr = = 0.306 sec
n
UnitstepResponse:
(d)FrequencydomainCharacteristics:
10 (1 + 5 s )
G(s) =
s
2
( s + 10 )
PM GM Mr BW
(deg) (dB) (rad/sec)
63 1.041 7.156
983
9-37) This problem is extensively discussed in Chapters 5 and 6. Use the transfer function (5-123) for
the open-loop system, and a series PID compensator in a unity feedback system.
Gc
Reduce to:
Gc G
in_Va
Where:
(s) Kt
G (s) = =
Va ( s ) s ( La Js + ( La B + Ra J ) s + Ra B + K t K b )
2
KI KDs + K ps + KI
2
Gc ( s ) = K p + K D s + =
s s
With
the rotor inertia (J) = 0.01 kg.m2/s2
damping ratio of the mechanical system (B) = 0.1 Nms
back-emf constant (Kb) = 0.01 Nm/Amp
torque constant (Kt) = 0.01 Nm/Amp
armature resistance (Ra) = 1
armature inductance (La) = 0.5 H
Starting systematically, set KI=KD=0. Assume a small electric time constant (or small inductance) and
simplify to Equation (5-126):
KK m K s
m ( s ) Ra
= (5-125)
in ( s ) KK KK K
( e s + 1) J m s 2 + B + b m s + m s
Ra Ra
984
WhereKsisthesensorgain,and,asbefore,e=(La/Ra)maybeneglectedforsmallLa.
KK t K s
m ( s ) Ra J
=
in ( s ) R B + Kt Kb KK t K s
s2 + a s+
Ra J Ra J
KK t K s
m ( s ) Ra J
=
in ( s ) R B + Kt Kb KK t K s
s2 + a s+ (5-126)
Ra J Ra J
K p 0.1
m ( s ) 0.01
=
in ( s ) 0.1 + (0.01)(0.01) K p 0.1
s2 + s+
0.01 0.01
m ( s ) 10 K p
=
in ( s ) s 2 + 12 s + 10 K p
Where Ks=0.
3 .2
Using ts ; for a less than 2 sec settling time n 1.6
n
For a PO of 4.3, =0.707, resulting in n=2.26.
Then a standard 2nd order prototype system that will have the desired response, with zero steady state
error, takes the following form
m ( s ) 5.12
= 2
in ( s ) s + 3.2 s + 5.12
For obvious reasons
m ( s ) 10 K p 5.12
= 2 2
in ( s ) s + 12 s + 10 K p s + 3.2 s + 5.12
Lets add a PD controller
Gc ( s ) = K p + K D s
985
m ( s ) 10 ( K D s + K p ) 5.12
= 2 2
in ( s ) s + (12 + 10 K D ) s + 10 K p s + 3.2 s + 5.12
(12 + 10 K D ) = 5.12
10 K p = 5.12
K D = 0.688
K p = 0.512
Although the two systems are not the same because one has a zero, we chose the controller gain values
by matching the two characteristic equations as an initial approximation. The resulting zero in the right
hand plane is troubling.
Lets find the response of the system through ACSYS:
Looking at the TF poles, it seems prudent to design the controller by placing its zero farther to LHS of
the s-plane. Set z=-Kp/KD=-20 and vary KD to find the root locus or the response.
986
Done!
9-38) The same as 9-37
987
9-39)
0 1 0 0 s 1 0 0
25.92 0 0 0 25.92 s 0 0
A = sI A =

0 0 0 1 0 0 s 1

2.36 0 0 0 2.36 0 0 s
s 0 0 25.92 0 0
= sI A = s 0 s 1 +

0 s 1 = s
2
(s 2
25.92 )
0 0 s 2.36 0 s
s3 s
2
0 0
2 3
1 25.92 s s 0 0
( sI A )
1
=
2.36 s 2.36 s 25.92 s s 25.92
3 2

2.36 s 2.36 s s 25.92 s
2 3
0
0 0.0732 s 2
0.0732 1 0.0732 s 3
( s I A )
1
B = ( s I A ) 0 =
1

0.0976 s 2.357
2

0.0976 0.0976 s 2.357 s
3
Y ( s ) = D ( sI A
0.0976 s 24.15 ( 2
)
) B = [0 (
0 ] sI A )
1 1
0 1 B=
s
2
(s 2
25.92 )
CharacteristicEquation: s + 0.0976 s + ( 0.0976 K P 25.92 ) s 2.357 K D s 2.357 K P = 0
4 3 2
3 1
ThesystemcannotbestabilizedbythePDcontroller,sincethe s andthe s termsinvolve K D which
requireoppositesignsfor K D .
988
9-40)
LetusfirstattempttocompensatethesystemwithaPIcontroller.
KI 100 ( K P s + K I )
Gc ( s ) = K P + Then G ( s ) = Gc ( s )G p ( s ) =
s (
s s + 10 s + 100
2
)
SincethesystemwiththePIcontrollerisnowatype1system,thesteadystateerrorofthesystemduetoa
stepinputwillbezeroaslongasthevaluesof K P and K I arechosensothatthesystemisstable.
Letuschoosetheramperrorconstant K v = 100 .Then, K I = 100. Thefollowingfrequencydomain
performancecharacteristicsareobtainedwith K I = 100 andvariousvalueof K P rangingfrom10to100.
KP PM GM Mr BW
10 1.60 29.70 50.13
20 6.76 7.62 69.90
30 7.15 7.41 85.40
40 6.90 8.28 98.50
50 6.56 8.45 106.56
100 5.18 11.04 160.00
ThemaximumphasemarginthatcanbeachievedwiththePIcontrollerisonly7.15degwhen K P = 30.
Thus,theovershootrequirementcannotbesatisfiedwiththePIcontrolleralone.
Next,wetryaPIDcontroller.
989
KI (1 + K s )( K s + KI 2 ) (1 + K s )( K s + 100 )
Gc ( s ) = K P + K D s + = D1 P2
= D1 P2

s s s
BasedonthePIcontrollerdesign,letusselect K P 2 = 30. Thentheforwardpathtransferfunction
becomes
100 ( 30 s + 100 )(1 + K D1 s )

G(s) =
(
s s + 10 s + 100
2
)
ThefollowingattributesofthefrequencydomainperformanceofthesystemwiththePIDcontroller
areobtainedforvariousvaluesof K D1rangingfrom0.05to0.4.
KD1 PM GM Mr BW
0.05 85.0 1.04 164.3
0.10 89.4 1.00 303.8
0.20 90.2 1.00 598.6
0.30 90.2 1.00 897.0
0.40 90.2 1.00 1201.0
Weseethatforvaluesof K D1greaterthan0.2,thephasemarginnolongerincreases,butthe
bandwidthincreaseswiththeincreasein K D1.Thuswechoose
K D1 = 0.2, K I = K I 2 = 100, K D = K D1 K P 2 = 0.2 30 = 6 ,
K P = K P 2 + K D1 K I 2 = 30 + 0.2 100 = 50
100
ThetransferfunctionofthePIDcontrolleris Gc ( s ) = 50 + 6 s +
s
Theunitstepresponseisshowbelow.Themaximumovershootiszero,andtherisetimeis0.0172sec.
990
991
9-41)
3.6 9 1 1
3.6 9
Lets consider the characteristic equation like:
3.6 9 2
for 2% settling time.
Therefore we can choose 0.5 and 4 where 2 3. Now, we can choose pole p
far enough from pole dominant of second order. Let p = 10, then the characteristic equation
would be:
36 9 14 56 160
where 160 , 0.29, and 0.065
160 0.065 0.29 1
Verify using MATLAB
992
9-42)
(a)
4 4 (1 + aTs ) 4 aTs
Gp (s) = G ( s ) = Gc ( s )G p ( s ) = Geq ( s ) =
s (1 + Ts )
3 2
Ts + s + 4
2 2
s
RootContours:(Tisfixedandavaries)
SelectasmallvalueforTandalargevaluefora.LetT=0.02anda=100.
1+ 2s 400 ( s + 0.5 )
Gc ( s ) = G(s) =
1 + 0.02 s s
2
( s + 50 )
3 2
Thecharacteristicequationis s + 50 s + 400 s + 200 = 0
Therootsare: 0.5355,9.3,40.17
Thesystemtransferfunctionis
Y (s) 400 ( s + 0.5 )

=
R(s) ( s + 0.5355 )( s + 9.3 )( s + 40.17 )
993
Sincethezeroat0.5isveryclosetothepoleat0.5355,thesystemcanbeapproximatedbyasecond
ordersystem,
Y (s) 373.48
=
R(s) ( s + 9.3 )( s + 40.17 )
Theunitstepresponseisshownbelow.Theattributesoftheresponseare:
Maximumovershoot=5% t s = 0.6225sec t r = 0.2173 sec
UnitstepResponse.
Thefollowingattributesofthefrequencydomainperformanceareobtainedforthesystemwiththe
phaseleadcontroller.
PM=77.4degGM=infiniteMr=1.05BW=9.976rad/sec
(b)TheBodeplotoftheuncompensatedforwardpathtransferfunctionisshownbelow.Thediagram
showsthattheuncompensatedsystemismarginallystable.ThephaseofG ( j ) is180degatall
994
frequencies.Forthephaseleadcontrollerweneedtoplace m atthenewgaincrossoverfrequency
torealizethedesiredphasemarginwhichhasatheoreticalmaximumof90deg.
Foradesiredphasemarginof80deg,
o
1 + sin 80
a= o
= 130
1 sin 80
Thegainofthecontrolleris 20 log10 a = 42 dB.Thenewgaincrossoverfrequencyisat
42
G ( j ) = = 21dB
2
4 2
Or G ( j ) = 2
= 0.0877 Thus = 45.61 = 6.75 rad/sec

1
= a m = 130 6.75 = 77 Thus T = 0.013
T
1
= 0.592 Thus aT = 169
.
aT
1 + aTs 1 + 1.702 s 4 (1 + 1.702 s )

Gc ( s ) = = G (s) =
1 + Ts 1 + 0.0131s s (1 + 0.0131s )
2
BodePlot.
995
996

1000a s +
1
1000 (1 + aTs )

G ( s ) = Gc ( s )GP ( s ) = =
aT

s ( s + 10 )(1 + Ts ) 1
s ( s + 10 ) s +
T
Set1/aT=10sothatthepoleofG ( s ) ats=10iscancelled.Thecharacteristicequationofthesystem
becomes
2 1
s + s + 1000 a = 0
T
1
n = 1000 a 2 n = = 2 1000 a Thusa=40andT=0.0025
T
ControllerTransferFunction: ForwardpathTransferFunction:
1 + 0.01s 40, 000

Gc ( s ) = G ( s ) =
1 + 0.0025 s s ( s + 400 )
Theattributesoftheunitstepresponseofthecompensatedsystemare:
Maximumovershoot=0 t r = 0.0168 sec t s = 0.02367 sec
(b)FrequencydomainDesign
TheBodeplotoftheuncompensatedforwardpathtransferfunctionismadebelow.
1000
G(s) = TheattributesofthesystemarePM=17.96deg,GM=infinite.
s ( s + 10 )
Mr=3.117,andBW=48.53rad/sec.
997
Torealizeaphasemarginof75deg,weneedmorethan57degofadditionalphase.Letusaddan
additional10degforsafety.Thus,thevalueof m forthephaseleadcontrollerischosentobe
67deg.Thevalueofaiscalculatedfrom
o
1 + sin 67
a= o
= 24.16
1 sin 67
Thegainofthecontrolleris 20 log10 a = 20 log10 24.16 = 27.66 dB.Thenewgaincrossoverfrequency
isat
' 27.66
G ( j m ) = = 13.83 dB
2
'
FromtheBodeplot m isfoundtobe70rad/sec.Thus,
1
= aT = 24.16 70 = 344 or T = 0.0029 aT = 0.0702
T
1 + aTs 1 + 0.0702 s
Thus Gc ( s ) = =
1 + Ts 1 + 0.0029 s
Thecompensatedsystemhasthefollowingfrequencydomainattributes:
PM=75.19deg GM=infinite Mr=1.024 BW=91.85rad/sec
Theattributesoftheunitstepresponseare:
Risetimetr=0.02278sec Settlingtimets=0.02828secMaximumovershoot=3.3%
998
999
9-44) Also see Chapter 6 for solution to this problem.
Mathematical Model:
Draw free body diagrams (Assume both xc and xw are positive and are measured from
equilibrium). Refer to Chapter 4 problems for derivation details.
kc ( xc xw ) cc ( x&c x&w )
mw mc

k w ( xw y ) cw ( x&w y& ) kc ( xc xw ) cc ( x&c x&w )
mw xw cw + cc
0 && cc x&w kw + kc kc xw cw 0 y& kw 0 y
0 + + = +
xc cc
mc && cc x&c kc kc xc 0 0 0 0 0 0
To Solve we need to simplify, since the problem is very difficult.
Assume the wheel is very stiff; hence k w = , which implies xw=y. Then
xc + cc x&c + kc xc = cc y& + kc y
mc &&
or
xc + 2n x&c + n2 xc = 2n y& + n2 y
&&
Placing an actuator between the two masses (ignore actuator dynamics for simplicity), and use a
PD control: the control force is (mc is added to make the final equation look simpler):
F = mc K D ( x&c x&w ) + mc K P ( xc xw )
where
xw = y, x&w = y&
The transfer function of the system is:
9100
Xc ( 2n + K D ) s + (n2 + K P )
= 2
Y s + ( 2n + K D ) s + ( 2n + K D )
The rest is a standard PD controller design.
9101
9-45)(a)ForwardpathTransferFunction:(N=10)
200 (1 + aTs )
G ( s ) = Gc ( s )G p ( s ) =
s ( s + 1)( s + 10 )(1 + Ts )
Startingwitha=1000,wevaryTfirsttostabilizethesystem.Thefollowingtimedomainattributes
areobtainedbyvaryingthevalueofT.
T MaxOvershoot tr ts
(%)
0.0001 59.4 0.370 5.205
0.0002 41.5 0.293 2.911
0.0003 29.9 0.315 1.83
0.0004 22.7 0.282 1.178
0.0005 18.5 0.254 1.013
0.0006 16.3 0.230 0.844
0.0007 15.4 0.210 0.699
0.0008 15.4 0.192 0.620
0.0009 15.5 0.182 0.533
0.0010 16.7 0.163 0.525
ThemaximumovershootisataminimumwhenT=0.0007orT=0.0008.Themaximumovershoot
is15.4%.
9102
UnitstepResponse.(T=0.0008seca=1000)
(b)FrequencydomainDesign.
Similartothedesigninpart(a),weseta=1000,andvarythevalueofTbetween0.0001and0.001.
Theattributesofthefrequencydomaincharacteristicsaregivenbelow.
PM GM Mr BW
T (deg) (dB) (rad/sec)
0.0001 17.95 60.00 3.194 4.849
0.0002 31.99 63.53 1.854 5.285
0.0003 42.77 58.62 1.448 5.941
0.0004 49.78 54.53 1.272 6.821
0.0005 53.39 51.16 1.183 7.817
0.0006 54.69 48.32 1.138 8.869
0.0007 54.62 45.87 1.121 9.913
0.0008 53.83 43.72 1.125 10.92
0.0009 52.68 41.81 1.140 11.88
0.0010 51.38 40.09 1.162 12.79
Thephasemarginisatamaximumof54.69degwhenT=0.0006.Theperformanceworsensifthevalueofaislessthan
1000.
9103
946(a)BodePlot.
Theattributesofthefrequencyresponseare:
PM=4.07deg GM=1.34dB Mr=23.24 BW=4.4rad/sec
(b)SinglestagePhaseleadController.
6 (1 + aTs )
G(s) =
s (1 + 0.2 s )(1 + 0.5 s )(1 + Ts )
Wefirstseta=1000,andvaryT.Thefollowingattributesofthefrequencydomaincharacteristics
areobtained.
T PM Mr
(deg)
0.0050 17.77 3.21
0.0010 43.70 1.34
0.0007 47.53 1.24
0.0006 48.27 1.22
0.0005 48.06 1.23
0.0004 46.01 1.29
0.0002 32.08 1.81
0.0001 19.57 2.97
Thephasemarginismaximumat48.27degwhenT=0.0006.
9104
Next,wesetT=0.0006andreduceafrom1000.Wecanshowthatthephasemarginisnotvery
sensitivetothevariationofawhenaisnear1000.Theoptimalvalueofaisaround980,andthe
correspondingphasemarginis48.34deg.
Witha=980andT=0.0006,theattributesoftheunitstepresponseare:
Maximumovershoot=18.8% tr=0.262sec ts=0.851sec
(c)TwostagePhaseleadController.(a=980,T=0.0006)
G(s) =
(
6 (1 + 0.588s ) 1 + bT s
2 )
( )

s (1 + 0.2 s )(1 + 0.5s )(1 + 0.0006 s ) 1 + T s
2
Again,letb=1000,andvary T2 .Thefollowingresultsareobtainedinthefrequencydomain.
PM Mr
T2
(deg)

0.0010 93.81 1.00
0.0009 94.89 1.00
0.0008 96.02 1.00
0.0007 97.21 1.00
0.0006 98.43 1.00
0.0005 99.61 1.00
0.0004 100.40 1.00
0.0003 99.34 1.00
0.0002 91.98 1.00
0.0001 73.86 1.00
Reducingthevalueofbfrom1000reducesthephasemargin.Thus,themaximumphasemarginof
100.4degisobtainedwithb=1000andT2=0.0004.Thetransferfunctionofthetwostagephase
leadcontrolleris
9105
(1 + 0.588s )(1 + 0.4 s )

Gc ( s ) =
(1 + 0.0006 s )(1 + 0.0004 s )
(c)UnitstepResponses.
9106
9-47) Also see derivations in 4-9.
x(t)
F
L
M
Here is an alternative representation including friction (damping) . In this case the angle is
measured differently.
Lets find the dynamic model of the system:
1)
2)
Let . If is small enough then 1 and , therefore
which gives:
s
F s
Ignoring friction 0.
s
F s
where
Ignoring actuator dynamics (DC motor equations), we can incorporate feedback control using a
series PD compensator and unity feedback. Hence,
9107
KP+KDs
R_F
F ( s) = K p ( R ( s ) ) K D s ( R ( s ) )
The system transfer function is:
A( K p + KDs)
= 2
R (s + K D s + A ( K p B )
Control is achieved by ensuring stability (Kp>B)
Use Routh Hurwitz to establish stability first. Use Acsys to do that as

demonstrated in this chapter problems. Also Chapter 2 has many examples.
Use MATLAB to simulate response:

clear all
Kp=10;
Kd=5;
A=10;
B=8;
num = [A*Kd A*Kp];
G=tf(num,den)
step(G)
Transfer function:
50 s + 100
--------------
s^2 + 5 s + 20
9108
Adjust parameters to achieve desired response. Use THE PROCEDURE in Example 5-11-1.
You may look at the root locus of the forward path transfer function to get a better perspective.
A ( K p + K D s ) AK D ( z + s )
= =
E s 2 AB s 2 AB
fix z and vary K D .
clear all
z=100;
Kd=0.01;
A=10;
B=8;
num = [A*Kd A*Kd*z];
den =[1 0 -(A*B)];
G=tf(num,den)
rlocus(G)
Transfer function:
0.1 s + 10
----------
s^2 80
9109
Forz=10,alargeKD=0.805resultsin:
clear all
Kd=0.805;
Kp=10*Kd;
A=10;
B=8;
num = [A*Kd A*Kp];
G=tf(num,den)
pole(G)
zero(G)
step(G)
Transfer function:
8.05 s + 80.5
-------------------
s^2 + 0.805 s + 0.5
ans =
9110
-0.4025 + 0.5814i
-0.4025 - 0.5814i
ans =
-10
Looking at dominant poles we expect to see an oscillatory response with overshoot

close to desired values.
For a better design, and to meet rise time criterion, use Example 5-
11-1 and Chapter 9 PD design examples.
9111
9-48)(a)Thelooptransferfunctionofthesystemis
6
10 K p K a K e 1 + R2 Cs 68.76 1 + R2 10 s
G(s) H (s) = =
Ns (1 + 0.05 s ) R1Cs s (1 + 0.05 s ) 2s
3 2 4
Thecharacteristicequationis s + 20 s + 6.876 10 R2 s + 687.6 = 0
Forrootlocusplotwith R2 asthevariableparameter,wehave
4 4
6.876 10 R2 s 6.876 10 R2 s
Geq ( s ) = =
s + 20 s + 687.6
3 2
( s + 21.5 )( s 0.745 + j 5.61)( s 0.745 j 5.61)
RootLocusPlot.
9112
5
When R2 = 2.65 10 ,therootsareat6.02 j 7.08 ,andtherelativedampingratiois0.65whichis
maximum.Theunitstepresponseisplottedattheendtogetherwiththoseofparts(b)and(c).
(b)PhaseleadController.
68.76 (1 + aTs )
G(s) H (s) =
s (1 + 0.05 s )(1 + Ts )
CharacteristicEquation: Ts + (1 + 20T ) s + ( 20 + 1375.2 aT ) s + 1375.2 = 0

3 2
WithT=0.01,thecharacteristicequationbecomes
3 2
s + 120 s + ( 2000 + 1375.2a ) s + 137520 = 0
Thelastequationisconditionedforarootcontourplotwithaasthevariableparameter.
Thus
1375.2as
Geq ( s ) = 3 2

s + 120 s + 2000 s + 137,520
Fromtherootcontourplotonthenextpageweseethatwhena=3.4thecharacteristicequationroots
areat39.2, 40.4 + j 43.3, and 40.4 j 43.3,andtherelativedampingratioismaximumandis
0.682.
RootContourPlot(avaries).
9113
UnitstepResponses.
9114
(c)FrequencydomainDesignofPhaseleadController.
Foraphasemarginof60deg,a=4.373andT=0.00923.Thetransferfunctionofthecontrolleris
1 + aTs 1 + 0.04036 s
Gc ( s ) = =
1 + Ts 1 + 0.00923 s
949(a)TimedomainDesignofPhaselagController.
ProcessTransferFunction:
200
Gp (s) =
s ( s + 1)( s + 10 )
Fortheuncompensatedsystem,thetwocomplexcharacteristicequationrootsareat s = 0.475 + j 0.471
and0.475 j 0.471whichcorrespondtoarelativedampingratioof0.707,whentheforwardpathgain
is4.5(asagainst200).Thus,thevalueofaofthephaselagcontrollerischosentobe
4.5
a= = 0.0225 SelectT=1000whichisalargenumber.
200
Then
1 + aTs 1 + 22.5 s 4.5 ( s + 0.0889) )

Gc ( s ) = = G ( s ) = Gc ( s )G p ( s ) =
1 + Ts 1 + 1000 s s ( s + 1)( s + 10 )( s + 0.001)
UnitstepResponse.
9115

Maximumovershoot=13.6 tr=3.238sec ts=18.86sec
BodePlot(withphaselagcontroller,a=0.0225,T=1000)
9116
PM=59deg. GM=27.34dB Mr=1.1BW=0.6414rad/sec
(b)FrequencydomainDesignofPhaselagController.
ForPM=60deg,wechoosea=0.02178andT=1130.55.Thetransferfunctionofthephaselag
controlleris
1 + 24.62 s
Gc ( s ) = GM=27.66dBMr=1.093BW=0.619rad/sec
1 + 1130.55s
9117
UnitstepResponse. Maxovershoot=12.6%,tr=3.297sects=18.18sec
950(a)TimedomainDesignofPhaseleadController
ForwardpathTransferFunction.
K (1 + aTs ) K
G ( s ) = Gc ( s )G p ( s ) = K v = lim sG ( s ) = = 10 ThusK=250
s ( s + 5 ) (1 + Ts )
2
s 0 25
WithK=250,thesystemwithoutcompensationismarginallystable.Fora>1,selectasmallvaluefor
Tandalargevaluefora.Leta=1000.ThefollowingresultsareobtainedforvariousvaluesofT
rangingfrom0.0001to0.001.WhenT=0.0004,themaximumovershootisnearminimumat23%.
T MaxOvershoot tr ts
(%) (sec) (sec)
0.0010 33.5 0.0905 0.808
0.0005 23.8 0.1295 0.6869
0.0004 23.0 0.1471 0.7711
9118
0.0003 24.4 0.1689 0.8765
0.0002 30.6 0.1981 1.096
0.0001 47.8 0.2326 2.399
Asitturnsouta=1000isnearoptimal.Ahigherorlowervalueforawillgivelargerovershoot.
UnitstepResponse.
(b)FrequencydomainDesignofPhaseleadController
250 (1 + aTs )
G(s) =
s
2
( s + 5 ) (1 + Ts )
2
Settinga=1000,andvaryingT,thefollowingattributesareobtained.
9119
T PM Mr BW
(deg) (rad/sec)
0.00050 41.15 1.418 16.05
0.00040 42.85 1.369 14.15
0.00035 43.30 1.355 13.16
0.00030 43.10 1.361 12.12
0.00020 38.60 1.513 10.04
Whena=1000,thebestvalueofTforamaximumphasemarginis0.00035,andPM=43.3deg.
Asitturnsoutvaryingthevalueofafrom1000doesnotimprovethephasemargin.Thusthe
transferfunctionofthecontrolleris
1 + aTs 1 + 0.35 s 250 (1 + 0.35 s )

Gc ( s ) = = and G(s) =
s ( s + 5 ) (1 + 0.00035s )
2
1 + Ts 1 + 0.00035 s
(c)TimedomainDesignofPhaselagController
Withoutcompensation,therelativedampingiscriticalwhenK=18.5.Then,thevalueofais
chosentobe
18.5
a = = 0.074
250
Wecanusethisvalueofaasareference,andconductthedesignaroundthispoint.Thevalueof
Tispreferrablytobelarge.However,ifTistoolarge,riseandsettlingtimeswillsuffer.
Thefollowingperformanceattributesoftheunitstepresponseareobtainedforvariousvaluesof
aandT.
9120
a T MaxOvershoot tr ts
(%)
0.105 500 2.6 1.272 1.82
0.100 500 2.9 1.348 1.82
0.095 500 2.6 1.422 1.82
0.090 500 2.5 1.522 2.02
0.090 600 2.1 1.532 2.02
0.090 700 1.9 1.538 2.02
0.090 800 1.7 1.543 2.02
0.090 1000 1.4 1.550 2.22
0.090 2000 0.9 1.560 2.22
0.090 3000 0.7 1.566 2.22
Asseenfromtheresults,whena=0.09andforT 2000,themaximumovershootisless
than1%andthesettlingtimeislessthan2.5sec.WechooseT=2000anda=0.09.
Thecorrespondingfrequencydomaincharacteristicsare:
PM=69.84deg GM=20.9dB Mr=1.004 BW=1.363rad/sec
(d)FrequencydomainDesignofPhaselagController
9121
250 (1 + aTs )
G ( s ) = a <1
s ( s + 5 ) (1 + Ts )
2
TheBodeplotoftheuncompensatedsystemisshownbelow.Letusaddasafetyfactorby
requiringthatthedesiredphasemarginis75degrees.Weseethataphasemarginof75degrees
canberealizedifthegaincrossoverismovedto0.64rad/sec.ThemagnitudeofG ( j ) atthis
frequencyis23.7dB.Thusthephaselagcontrollermustprovideanattenuationof23.7dBat
thenewgaincrossoverfrequency.Setting
20 log10 a = 23.7 dB wehavea=0.065
Wecansetthevalueof1/aTtobeatleastonedecadebelow0.64rad/sec,or0.064rad/sec.Thus,
wegetT=236.LetuschooseT=300.Thetransferfunctionofthephaselagcontrollerbecomes
1 + aTs 1 + 19.5s
Gc ( s ) = =
1 + Ts 1 + 300 s
9122
9123
Theattributesofthefrequencyresponseofthecompensatedsystemare:
PM=71deg GM=23.6dB Mr=1.065 BW=0.937rad/sec
Maximumovershoot=6%tr=2.437sec ts=11.11sec
Comparingwiththephaselagcontrollerdesignedinpart(a)whichhasa=0.09andT=2000,
thetimeresponseattributesare:
Maximumovershoot=0.9% tr=1.56sec ts=2.22sec
ThemaindifferenceisinthelargevalueofTusedinpart(c)whichresultedinlessovershoot,
riseandsettlingtimes.
9124
9-51)
9-52)ForwardpathTransferFunction(Nocompensation)
6.087 10
7
G(s) = Gp (s) =
(
s s + 423.42 s + 2.6667 10 s + 4.2342 10
3 2 6 8
)
Theuncompensatedsystemhasamaximumovershootof14.6%.Theunitstepresponseisshown
below.
(a)PhaseleadController
1 + aTs
Gc ( s ) = (a>1)
1 + Ts
ByselectingasmallvalueforT,thevalueofabecomesthecriticaldesignparameterinthiscase.
Ifaistoosmall,theovershootwillbeexcessive.Ifthevalueofaistoolarge,theoscillationin
thestepresponsewillbeobjectionable.Bytrialanderror,thebestvalueofaisselectedtobe6,
andT=0.001.Thefollowingperformanceattributesareobtainedfortheunitstepresponse.
Maximumovershoot=0%tr=0.01262sec ts=0.1818sec
However,thestepresponsestillhasoscillationsduetothecomplianceinthemotorshaft.Theunit
stepresponseofthephaseleadcompensatedsystemisshownbelow,togetherwiththatofthe
uncompensatedsystem.
(b)PhaseleadandSecondorderController
9125
Thepolesoftheprocess G p ( s ) areat161.3,131+j1614.7and131j1614.7.Thesecond
2
ordertermis s + 262 s + 2,624, 417.1.Letthesecondordercontrollertransferfunctionbe
2
s + 262 s + 2,624,417.1
Gc1 ( s ) = 2 2

s + 2 p n s + n
Thevalueof n issetto 2,624, 417.1 = 1620 rad/sec,sothatthesteadystateerrorisnotaffected.
LetthetwopolesofGc1 ( s ) beat s = 1620 and 1620. Then, p = 405 .
2
s + 262 s + 2,624,417.1
Gc1 ( s ) = 2

s + 3240 s + 2,624,417.1
6.087 10
10
(1 + 0.006 s )
G ( s ) = Gc ( s )Gc1 ( s )G p ( s ) =
( )
s ( s + 161.3 ) s + 3240 s + 2, 624, 417.1 (1 + 0.001s )
2
Theunitstepresponseisshownbelow,andtheattributesare:
Maximumovershoot=0.2tr=0.01012sec ts=0.01414sec
Thestepresponsedoesnothaveanyripples.
UnitstepResponses
9126
953(a)SystemEquations.
d m
ea = Ra ia + Kb m Tm = Ki ia Tm = J m + Bm m + K L ( m L ) + BL ( m L )
dt
d L
K L ( m L ) + BL ( m L ) = J L
dt
StateEquationsinVectormatrixForm:
9127
d L
dt 0 1 0 0
K BL KL BL 0
d L L L 0
dt J L JL JL JL L
d = 0 + 0 ea
m K
0 0 1
m
B + Bl K i K b m
a
dt K L BL KL
d m J m
a m
R J
m Jm Jm Jm J m Ra
dt
StateDiagram:
TransferFunctions:
m (s)
=
( )
K i s + BL s + K L / Ra
2

J m J L s + ( K e J L + BL J L + BL J m ) s + ( J m K L + J L K L + K e BL ) s + K L K e
2 2
Ea ( s )
L (s) K i ( BL s + K L ) / Ra
=
J m J L s + ( K e J L + B L J L + B L J m ) s + ( J m K L + J L K L + K e BL ) s + K L K e
3 2
Ea ( s )
9128
m (s)
=
(
133.33 s + 10 s + 3000
2
) =
(
133.33 s + 10 s + 3000
2
)
Ea ( s ) s + 318.15 s + 60694.13s + 58240
3 2
( s + 0.9644 )( s + 158.59 + j187.71)( s + 158.59 j187.71)
L (s) 1333.33 ( s + 300 )

=
Ea ( s ) ( s + 0.9644 )( s + 158.59 + j187.71)( s + 158.59 j187.71)
(b)DesignofPIController.
KI
L (s)
1333.33 K P s + ( s + 300 )
G ( s ) = =
KP

E (s) s ( s + 0.9644 ) ( s + 317.186 s + 60388.23
2
)
1333.33 300 K I
K v = lim sG ( s ) = = 6.87 K I = 100 Thus K I = 14.56
s 0 0.9644 60388.23
With K I = 14.56 ,westudytheeffectsofvarying K P .Thefollowingresultsareobtained.
KP tr ts MaxOvershoot
(sec) (sec) (%)
20 0.00932 0.02778 4.2
18 0.01041 0.01263 0.7
17 0.01113 0.01515 0
16 0.01184 0.01515 0
15 0.01303 0.01768 0
10 0.02756 0.04040 0.6
With K I = 14.56 and K P rangingfrom15to17,thedesignspecificationsaresatisfied.
9129
UnitstepResponse:
(c)FrequencydomainDesignofPIController(KI=14.56)
1333.33 ( K P s + 14.56 )( s + 300 )

G(s) =
(
s s + 318.15s + 60694.13s + 58240
3 2
)
9130
Thefollowingresultsareobtainedbysetting K I = 14.56 andvaryingthevalueof K P .
KP PM GM Mr BW MaxOvershoot tr ts
(deg) (dB) (rad/sec) (%) (sec) (sec)
20 65.93 1.000 266.1 4.2 0.00932 0.02778
18 69.76 1.000 243 0.7 0.01041 0.01263
17 71.54 1.000 229 0 0.01113 0.01515
16 73.26 1.000 211.6 0 0.01184 0.01515
15 74.89 1.000 190.3 0 0.01313 0.01768
10 81.11 1.005 84.92 0.6 0.0294 0.0404
8 82.66 1.012 63.33 1.3 0.04848 0.03492
7 83.14 1.017 54.19 1.9 0.03952 0.05253
6 83.29 1.025 45.81 2.7 0.04697 0.0606
5 82.88 1.038 38.12 4.1 0.05457 0.0606
Fromtheseresultsweseethatthephasemarginisatamaximumof83.29degreeswhen K P = 6 .
However,themaximumovershootoftheunitstepresponseis2.7%,and M r isslightlygreaterthan
one.Inpart(b),theoptimalvalueof K P fromthestandpointofminimumvalueofthemaximum
overshootisbetween15and17.Thus,thephasemargincriterionisnotagoodindicatorinthepresent
case.
9131
954(a)ForwardpathTransferFunction
G p ( s ) =
K m (s)
=
(
100 K s + 10 s + 100
2
=
) ( 10, 000 s + 10 s + 100
2
)
Tm ( s ) (
s s + 20 s
3 2
+ 2100 s + 10, 000 ) s ( s + 4.937 ) ( s
2
+ 15.06 s + 2025.6 )
Theunitstepresponseisplottedasshownbelow.Theattributesoftheresponseare:
Maximumovershoot=57% tr=0.01345sec ts=0.4949sec
(b)DesignoftheSecondorderNotchController
Thecomplexzerosofthenotchcontrolleraretocancelthecomplexpolesoftheprocesstransfer
function.Thus
Gc ( s ) =
2
s + 15.06 s + 2025.6
and G ( s ) = G ( s )G ( s ) =
(
10, 000 s + 10 s + 100
2
)
2
s ( s + 4.937 ) ( s )
c p
s + 90 p s + 2025.6 + 90 p s + 2025.6
2
Thefollowingresultsareobtainedfortheunitstepresponsewhenvariousvaluesof p areused.
Themaximumovershootisataminimumof4.1%when p = 1.222 .Theunitstepresponseis
plottedbelow,alongwiththatoftheuncompensatedsystem.
p 2 n MaxOvershoot
(%)
2.444 200 7.3
2.333 210 6.9
2.222 200 6.5
9132
1.667 150 4.9
1.333 120 4.3
1.222 110 4.1
1.111 100 5.8
1.000 90 9.8
UnitstepResponse
(c)FrequencydomainDesignoftheNotchController
Theforwardpathtransferfunctionoftheuncompensatedsystemis
9133
G ( s) =
(
10000 s + 10 s + 100
2
)
s ( s + 4.937 ) ( s + 15.06 s + 2025.6 )
2
TheBodeplotofG ( j ) isconstructedinthefollowing.Weseethatthepeakvalueof G ( j ) is
approximately22dB.Thus,thenotchcontrollershouldprovideanattentuationof22dBor0.0794
attheresonantfrequencyof45rad/sec.UsingEq.(10155),wehave
z 0.167
Gc ( j 45) = = = 0.0794 Thus p = 2.1024
p p
NotchControllerTransferFunction ForwardpathTransferFunction
Gc ( s ) =
2
s + 15.06 s + 2025.6
G ( s ) =
(
10, 000 s + 10 s + 100
2
)
s ( s + 4.937 ) ( s )
2
s + 189.216 s + 2025.6 + 189.22 s + 2025.6
2
BodePlots
9134
Attributesofthefrequencyresponse:PM=80.37degGM=infinite M r = 1097
. BW=66.4rad/sec
Attributesofthefrequencyresponseofthesystemdesignedinpart(b):
PM=59.64deg GM=infinite M r = 1048

. BW=126.5rad/sec
9135
955(a)ProcessTransferFunction
500 ( s + 10 )
Gp ( s) =
(
s s + 10 s + 1000
2
)
TheBodeplotisconstructedbelow.Thefrequencydomainattributesoftheuncompensated
systemare:
PM=30deg GM=infinite M r = 186

. andBW=3.95rad/sec
Theunitstepresponseisoscillatory.
(b)DesignoftheNotchController
Fortheuncompensatedprocess,thecomplexpoleshavethefollowingconstants:
n = 1000 = 316
. rad / sec 2 n = 10 Thus = 0.158
Thetransferfunctionofthenotchcontrolleris
2 2
s + 2 z n s + n
Gc ( s ) = 2 2

s + 2 p s + n
Forthezerosof Gc ( s ) tocancelthecomplexpolesof G p ( s ) , z = = 0.158 .
FromtheBodeplot,weseethattobringdownthepeakresonanceof G ( j n ) inorder
tosmoothoutthemagnitudecurve,thenotchcontrollershouldprovideapproximately
9136
26dBofattenuation.Thus,usingEq.(10155),
26
z 0.158
= 10 20
= 0.05 Thus p = = 31525
.
p 0.05
Thetransferfunctionofthenotchcontrolleris
2
s + 10 s + 1000 500 ( s + 10 )
Gc ( s ) = G ( s ) = Gc ( s )G p ( s ) =
2
s + 199.08 s + 1000 (
s s + 199.08s + 1000
2
)
Theattributesofthecompensatedsystemare:
PM=72.38deg GM=infinite M r = 1 BW=5.44rad/sec
Maximumovershoot=3.4% t r = 0.3868 sec t s = 0.4848 sec
BodePlots
9137
StepResponses
9138
(c)TimedomaindesignoftheNotchController
With z = 0.158 and n = 316

. ,theforwardpathtransferfunctionofthecompensatedsystemis
500 ( s + 10 )
G ( s ) = Gc ( s )G p ( s ) =
(
s s + 63.2 p s + 1000
2
)
Thefollowingattributesoftheunitstepresponseareobtainedbyvaryingthevalueof p .
p 2 n MaxOvershoot tr ts
(%) (sec) (sec)
1.582 100 0 0.4292 0.5859
1.741 110 0 0.4172 0.5657
1.899 120 0 0.4074 0.5455
2.057 130 0 0.3998 0.5253
2.215 140 0.2 0.3941 0.5152
2.500 158.25 0.9 0.3879 0.4840
3.318 209.7 4.1 0.3884 0.4848
When p = 2.5 themaximumovershootis0.9%,therisetimeis0.3879secandthesetting
timeis0.4840sec.Theseperformanceattributesarewithintherequiredspecifications.
9139
956Letthetransferfunctionofthecontrollerbe
Gc ( s ) =
(
20, 000 s + 10 s + 50
2
)
( s + 1000 ) 2
Then,theforwardpathtransferfunctionbecomes
G ( s ) = Gc ( s )G p ( s ) =
(
20, 000 K s + 10 s + 50
2
)
(
s s + 10 s + 100
2
) ( s + 1000 ) 2
6
10 K
For Gcf ( s ) = 1, K v = lim sG ( s ) = 8
= 50 Thus the nominal K = 5000
s 0 10
For 20%variationinK, K min = 4000 and K max = 6000. Tocancelthecomplexclosedlooppoles,
welet
50 ( s + 1)
Gcf ( s ) = wherethe(s+1)termisaddedtoreducetherisetime.
s + 10 s + 50
2
ClosedloopTransferFunction:
10 K ( s + 1)
6
Y (s)
=
R(s) (
s s + 10 s + 100
2
)( s + 1000 ) + 20, 000 K ( s
2 2
+ 10 s + 50 )

5 4 3 7 2 8 9
K=4000: s + 2010 s + 1,020,100 s + 9.02 10 s + 9 10 + 4 10 = 0
Roots: 97.7, 648.9, 1252.7, 5.35 + j 4.6635, 5.35 j 4.6635
Maxovershoot 6.7% Risetime<0.04sec
9140
5 4 3 8 2 9 9
K=5000: s + 2010 s + 1,020100 s + 11
. 10 s + 11
. 10 s + 5 10 = 0
Roots: 132.46, 587.44, 1279.6, 5.272 + j 4.7353, 5.272 j 4.7353
Maxovershoot 4% Risetime<0.04sec
5 4 3 8 2 9 9
K=6000 s + 2010 s + 1,020,100 s + 13
. 10 s + 13
. 10 s + 6 10 = 0
Roots: 176.77, 519.37, 1303.4, 5.223 + j 4.7818, 5.223 j 4.7818
Maxovershoot 2.5% Risetime<0.04sec
ThusalltherequiredspecificationsstaywithintherequiredtoleranceswhenthevalueofKvariesby
plusandminus20%.
UnitstepResponses
9141
957Letthetransferfunctionofthecontrollerbe
Gc ( s ) =
(
200 s + 10 s + 50
2
)
( s + 100 ) 2
G ( s ) = Gc ( s )G p ( s ) =
(
200, 000 K s + 10 s + 50
2
)
s ( s + a )( s + 100 )
2
Fora=10,
7
10 K
K v = lim sG ( s ) = 5
= 100 K = 100 Thus K = 1
s 0 10
CharacteristicEquations:(K=1)
4 3 5 2 6 7
a=10: s + 210 s + 2.12 10 s + 2.1 10 s + 10 = 0
Roots: 4.978 + j 4.78, 4.978 j 4.78, 100 + j 447.16 100 j 447.16
4 3 5 2 6 7
a=8: s + 208 s + 2.116 10 s + 2.08 10 s + 10 = 0
Roots: 4.939 + j 4.828, 4.939 j 4.828, 99.06 + j 446.97, 99.06 j 446.97
4 3 5 6 7
a=12: s + 212 s + 2.124 10 + 2.12 10 s + 10 = 0
Roots: 5.017 + j 4.73, 5.017 j 4.73, 100.98 + j 447.36, 100.98 j 447.36
9142
UnitstepResponses:Allthreeresponsesfora=8,a=10,and12aresimilar.
Y ( s) K K
G (s) = = K v = lim sG ( s ) = =1
E (s) s ( s + 1)( s + 10 ) + KK t s s 0 10 + KK t
CharacteristicEquation: s + 11s + (10 + KK t ) s + K = s + 11s + Ks + K = 0

3 2 3 2
K ( s + 1)
Forrootloci, Geq ( s ) =
s
2
( s + 11)
9143
RootLocusPlot(Kvaries)
Therootlocishowthatarelativedampingratioof0.707canberealizedbytwovaluesofK.K=22
and59.3.Asstipulatedbytheproblem,weselectK=59.3.
9144
10 K 10 K K
G(s) = K v = lim sG ( s ) = = = 1 Thus Kt = K 1
s ( s + 1)( s + 10 ) + 10 K t s s 0 10 + 10 Kt 1 + Kt
CharacteristicEquation: s ( s + 1)( s + 10 ) + 10 K t + 10 K = s + 11s + 10 Ks + 10 K = 0

3 2
WhenK=5.93and Kt = K 1 = 4.93 ,thecharacteristicequationbecomes
3 2
s + 11s + 10.046 s + 4.6 = 0
Therootsare:10.046, 0.47723 + j 0.47976, 0.47723 j 0.47976
K (1 + aTs ) 1
G(s) = K v = lim sG ( s ) = = 100 Thus Kt = 0.01
(
s ( (1 + Ts ) s + 10 s + KK t
2
) s 0 Kt
LetT=0.01anda=100.Thecharacteristicequationofthesystemiswritten:

2
(
s + 110 s + 1000 s + K 0.001s + 101s + 100 = 0
4 3 2
)
ToconstructtherootcontoursasKvaries,weformthefollowingequivalentforwardpathtransfer
function:
Geq ( s ) =
(
0.001K s + 101, 000 s + 100, 000
2
) = 0.001K ( s + 1)( s + 50499 )
s
2
( s + 10 )( s + 100 ) s
2
( s + 10 )( s + 100 )
9145
Fromtherootcontourdiagramweseethattwosetsofsolutionsexistforadampingratioof0.707.
Theseare:
K=20: Complexroots: 1158

. + j1155
. , 1158
. j1155
.
K=44.6:Complexroots:4.0957 + j 4.0957, 4.0957 j 4.0957
TheunitstepresponsesofthesystemforK=20and44.6areshownbelow.
UnitstepResponses:
9146
K s KK1 K i N
G(s) =
s J t La s + ( Ra J t + La Bt + K1 K 2 J t ) s + Ra Bt + K1 K 2 Bt + K b K i + KK1 K i K t
2
1.5 10 K
7
G ( s) =
(
s s + 3408.33s + 1, 204, 000 + 1.5 10 KK t
2 8
)
15 K
RampErrorConstant: K v = lim sG ( s ) = = 100
s 0 1.204 + 150 KKt
Thus 1.204 + 150 KKt = 0.15 K
9147
1.5 10 K
7
G(s) =
(
s s + 3408.33s + 150, 000 K
2
)
3 7
CharacteristicEquation: s + 3408.33 s + 150,000 Ks + 15
. 10 K = 0
WhenK=38.667therootsofthecharacteristicequationareat
0.1065, 1651
. + j165
. , 1651
. j165
. ( 0.707 forthecomplexroots)
5.8 10
8
G(s) =
(
s s + 3408.33s + 5.8 10
2 6
)
UnitstepResponse
9148
Unitstepresponseattributes:Maximumovershoot=0Risetime=0.0208secSettlingtime=0.0283sec
962 (a)DisturburncetoOutputTransferFunction
Y (s) 2 (1 + 0.1s )
== Gc ( s ) = 1
TL ( s ) r =0
s (1 + 0.01s )(1 + 0.1s ) + 20 K
For TL ( s ) = 1 / s
1
lim y ( t ) = lim sY ( s ) = 0.01 Thus K 10
t s 0 10 K
(b)PerformanceofUncompensatedSystem.K=10,Gc ( s ) = 1
9149
200
G(s) =
s (1 + 0.01s )(1 + 0.1s )
TheBodediagramof G ( j ) isshownbelow.Thesystemisunstable.Theattributesofthe
frequencyresponseare:PM=9.65degGM=5.19dB.
(c)SinglestagePhaseleadControllerDesign
Torealizeaphasemarginof30degrees,a=14andT=0.00348.
1 + aTs 1 + 0.0487 s
Gc ( s ) = =
1 + Ts 1 + 0.00348 s
TheBodediagramofthephaseleadcompensatedsystemisshownbelow.Theperformance
attributesare:PM=30deg GM=10.66dB M r = 1.95 BW=131.6rad/sec.
(d)TwostagePhaseleadControllerDesign
200 (1 + 0.0487 s )
Startingwiththeforwardpathtransferfunction G ( s ) =
s (1 + 0.1s )(1 + 0.01s )(1 + 0.00348 s )
Theproblembecomesthatofdesigningasinglestagephaseleadcontroller.Foraphasemargin
or55degrees,a=7.385andT=0.00263.Thetransferfunctionofthesecondstagecontrolleris
1 + aTs 1 + 0.01845 s
Gc1 ( s ) = =
1 + Ts 1 + 0.00263 s
9150
200 (1 + 0.0487 s )(1 + 0.01845 s )

Thus G(s) =
s (1 + 0.1s )(1 + 0.01s )(1 + 0.00348 s )(1 + 0.00263s )
TheBodediagramisshownonthefollowingpage.Thefollowingfrequencyresponseattributes
areobtained:
PM=55deg GM=12.94dB M r = 111

. BW=256.57rad/sec
BodePlot[parts(b),(c),and(d)]
9151
9152
963(a)TwostagePhaseleadControllerDesign.
Theuncompensatedsystemisunstable.PM=43.25degandGM=18.66dB.
Withasinglestagephaseleadcontroller,themaximumphasemarginthatcanberealizedaffectively
is12degrees.SettingthedesiredPMat11deg,wehavetheparametersofthesinglestagephase
leadcontrollerasa=128.2andT1 = 0.00472 .Thetransferfunctionofthesinglestagecontroller
1 + aT1 s 1 + 0.6057 s
is Gc1 ( s ) = =
1 + T1 s 1 + 0.00472 s
Startingwiththesinglestagecontrollercompensatedsystem,thesecondstageofthephaselead
controllerisdesignedtorealizeaphasemarginof60degrees.Theparametersofthesecondstage
controllerare:b=16.1andT2 = 0.0066 .Thus,
1 + bT2 s 1 + 0.106 s
Gc 2 ( s ) = =
1 + T2 s 1 + 0.0066 s
1 + 0.6057 s 1 + 0.106 s
Gc ( s ) = Gc1 ( s )Gc 2 ( s ) =
1 + 0.00472 s 1 + 0.0066 s
1, 236, 598.6 ( s + 1.651)( s + 9.39 )

G ( s ) = Gc1 ( s )Gc 2 ( s )G p ( s ) =
s ( s + 2)( s + 5 )( s + 211.86 )( s + 151.5 )
Attributesofthefrequencyresponseofthecompensatedsystemare:
GM=19.1dB PM=60deg M r = 1.08 BW=65.11rad/sec
Theunitstepresponseisplottedbelow.Thetimeresponseattributesare:
Maximumovershoot=10.2% t s = 0.1212 sec t r = 0.037 sec
9153
(b)SinglestagePhaselagControllerDesign.
Withasinglestagephaselagcontroller,foraphasemarginof60degrees,a=0.0108andT=1483.8.
Thecontrollertransferfunctionis
1 + 16.08 s
Gc ( s ) =
1 + 1483.8 s
Theforwardpathtransferfunctionis
G ( s ) = Gc ( s )G p ( s ) =

6.5 s + 0.0662
s
s + 2
s + 5s + 0.000674

FromtheBodeplot,thefollowingfrequencyresponseattributesareobtained:
PM=60deg GM=20.27dB M r = 1.09 BW=1.07rad/sec
Theunitstepresponsehasalongrisetimeandsettlingtime.Theattributesare:
Maximumovershoot=12.5% t s = 12.6 sec t r = 2.126 sec
(c)LeadlagControllerDesign.
Fortheleadlagcontroller,wefirstdesignthephaselagportionfora40degreephasemargin.
Theresultisa=0.0238andT1 = 350 .Thetransferfunctionofthecontrolleris
1 + 8.333 s
Gc1 ( s ) =
1 + 350 s
Thephaseleadportionisdesignedtoyieldatotalphasemarginof60degrees.Theresultis
b=4.8and T2 = 0.2245.Thetransferfunctionofthephaseleadcontrolleris
1 + 1076
. s
Gc 2 ( s ) =
1 + 0.2245s
9154
Theforwardpathtransferfunctionoftheleadlagcompensatedsystemis
68.63 ( s + 0.12 )( s + 0.929 )

G(s) =
s ( s + 2 )( s + 5 )( s + 0.00286 )( s + 4.454 )
Frequencyresponseattributes:PM=60degGM=13.07dB M r = 1.05 BW=3.83rad/sec
Unitstepresponseattributes:Maximumovershoot=5.9% t s = 1512
. sec t r = 0.7882 sec
UnitstepResponses.
964(a)Theuncompensatedsystemhasthefollowingfrequencydomainattributes:
PM=3.87deg GM=1dB M r = 7.73BW=4.35rad/sec
TheBodeplotof G p ( j ) showsthatthephasecurvedropsoffsharply,sothatthephaselead
controllerwouldnotbeveryeffective.Considerasinglestagephaselagcontroller.Thephase
9155
marginof60degreesisrealizedifthegaincrossoverismovedfrom2.8rad/secto0.8rad/sec.
Theattenuationofthephaselagcontrollerathighfrequenciesisapproximately15dB.
Choosinganattenuationof17.5dB,wecalculatethevalueofafrom
20 log10 a = 17.5 dBThus a=0.1334
Theuppercornerfrequencyofthephaselagcontrollerischosentobeat1/aT=0.064rad/sec.
Thus,1/T=0.00854orT=117.13.Thetransferfunctionofthephaselagcontrolleris
1 + 15.63 s
Gc ( s ) =
1 + 117.13
Theforwardpathtransferfunctionis
5 (1 + 15.63s )(1 0.05 s )

G ( s ) = Gc ( s ) G p ( s ) =
s (1 + 0.1s )(1 + 0.5 s )(1 + 117.13s )(1 + 0.05 s )
FromtheBodeplotof G ( j ) ,thefollowingfrequencydomainattributesareobtained:
PM=60deg GM=18.2dB M r = 1.08 BW=1.13rad/sec
Theunitstepresponseattributesare:
maximumovershoot=10.7% t s = 10.1 sec t r = 2.186 sec
BodePlots
9156
9157
964(b)Usingtheexactexpressionofthetimedelay,thesamedesignholds.Thetimeandfrequencydomain
attributesarenotmuchaffected.
965(a)UncompensatedSystem.
10
ForwardpathTransferFunction: G(s) =
(1 + s )(1 + 10 s )(1 + 2 s )(1 + 5s )
TheBodeplotof G ( j ) isshownbelow.
Theperformanceattributesare:PM=10.64degGM=2.26dB
Theuncompensatedsystemisunstable.
(b)PIControllerDesign.
10 ( K p s + K I )
ForwardpathTransferFunction: G (s) =
s (1 + s )(1 + 10 s )(1 + 2 s )(1 + 5 s )
RamperrorConstant: K v = lim sG ( s ) = 10 K I = 0.1 Thus K I = 0.01

s 0
0.1 (1 + 100 K P s )
G ( s) =
s (1 + s )(1 + 10 s )(1 + 2 s )(1 + 5 s )
Thefollowingfrequencydomainattributesareobtainedforvariousvaluesof K P .
KP PM GM Mr BW
0.01 24.5 5.92 2.54 0.13
0.02 28.24 7.43 2.15 0.13
0.05 38.84 11.76 1.52 0.14
9158
0.10 50.63 12.80 1.17 0.17
0.12 52.87 12.23 1.13 0.18
0.15 53.28 11.22 1.14 0.21
0.16 52.83 10.88 1.16 0.22
0.17 51.75 10.38 1.18 0.23
0.20 49.08 9.58 1.29 0.25
Thephasemarginismaximumat53.28degreeswhen K P = 0.15.
Theforwardpathtransferfunctionofthecompensatedsystemis
0.1 (1 + 15 s )
G ( s) =
s (1 + s )(1 + 10 s )(1 + 5 s )(1 + 2 s )
Theattributesofthefrequencyresponseare:
PM=53.28deg GM=11.22dB M r = 114

. BW=0.21rad/sec
Maximumovershoot=14.1% t r = 10.68 sec t s = 48 sec
BodePlots
9159
9160
965(c)TimedomainDesignofPIController.
Bysetting K I = 0.01andvarying K P wefoundthatthevalueof K P thatminimizesthemaximum
overshootoftheunitstepresponseisalso0.15.Thus,theunitstepresponseobtainedinpart(b)
isstillapplicableforthiscase.
966ClosedloopSystemTransferFunction.
Y (s) 1
=
s + ( 4 + k3 ) s + ( 3 + k 2 + k3 ) s + k1
3 2
R(s)
Forzerosteadystateerrortoastepinput, k1 = 1.Forthecomplexrootstobelocatedat1+jand1j,
2
wedividethecharacteristicpolynomialby s + 2 s + 2 andsolveforzeroremainder.
s + ( 2 + k2 )
s + 2 s + 2 s + ( 4 + k3 ) s + ( 3 + k 2 + k3 ) s + 1
2 3 2

s + + 2s
3 2
2s
(2 + k ) s
3
2
+ (1 + k 2 + k 3 ) s + 1
(2 + k ) s
3
2
+ ( 4 + 2 k3 ) s + 4 + 2 k3

( -3+k 2
k3 ) s 3 2k3
Forzeroremainder, 3 2 k3 = 0 Thus k3 = 15
.
3 + k 2 k3 = 0 Thus k 2 = 15
.
Thethirdrootisat0.5.Notalltherootscanbearbitrarilyassigned,duetotherequirementonthe
steadystateerror.
9161
967(a)OpenloopTransferFunction.
X 1 (s) k3
G ( s) = =
s s + ( 4 + k 2 ) s + 3 + k1 + k 2
2
E (s)
Sincethesystemistype1,thesteadystateerrorduetoastepinputiszeroforallvaluesof k1 , k 2 , and k3
thatcorrespondtoastablesystem.Thecharacteristicequationoftheclosedloopsystemis
s + ( 4 + k 2 ) s + ( 3 + k1 + k 2 ) s + k 3 = 0
3 2

Fortherootstobeat1+j,1j,and10,theequationshouldbe:
3 2
s + 12 s + 22 s + 20 = 0
Equatinglikecoefficientsofthelasttwoequations,wehave
4 + k 2 = 12 Thus k2 = 8
3 + k1 + k 2 = 22 Thus k1 = 11
k3 = 20 Thusk3=20
(b)OpenloopTransferFunction.
Y (s) Gc ( s ) 20 20 ( s + 1)( s + 3 )
= = Thus Gc ( s ) =
E ( s) ( s + 1)( s + 3 ) (
s s + 12 s + 22
2
) (
s s + 12 s + 22
2
)
9162
Chapter 10_________________________________________________________________________
dy
10-1)(a)Statevariables: x1 = y, x 2 =
dt
Stateequations: Outputequation:
dx1
dt 0 1 x1 0 x1
= x + 5 r y = [1 0] = x1
2
dx 1 4 2 x2
dt
2
dy d y
(b)Statevariables: x1 = y, x 2 = , x3 = 2

dt dt
dx1
dt
0 1 0 x1 0 x1

= 0 1 x2 + 0 r y = [1 0 0 ] x2 = x1
dx 2
0
dt
dx 1 2.5 1.5 x3 0.5 x3
3
dt
2
dx1 dy d y

t
(c)Statevariables: x1 = y ( ) d , x2 = , x3 = , x4 = 2

0
dt dt dt
x&1 0 1 0 0 x1 0 x1
x& 0 0 1 0 x 0 x
= 2 + r y = [1 0 0 0 ] = x1
2 2
x&3 0 0 0 1 x3 0 x3
&
x4 1 1 3 5 x4 1 x4
2 3
dy d y d y
(d)Statevariables: x1 = y, x 2 = , x3 = 2
, x4 = 3

dt dt dt
x&1 0 1 0 0 x1
0 x1
x& 0 0 1 0 x 0 x
= 2+ r y = [1 0 0 0 ] = x1
2 2

x&3 0 0 0 1 x3 0 x3
&
x4 1 2.5 0 1.5 x4 1 x4
101
10-2) a)
3 2 3
3
Let
3 3
Then
2 3 2 3
If and , then
3
2 3
or
3 1 1
2 0 3
1 0
b)
6 11 6 6
6
Let X s , therefore 11
and Let , then 6 6 . As a result:
6
11
6 6
or
6
11
6 6
If , , then
6 1 0 0
11 0 1 0 u
6 0 0 6
102
1 0 0
c)
7 12 2
7
Let , then 12 2 . As a result:
7
12 2
Let and , then
7 1 1
12 0 2
1 0
d)
4 39 108 11
Let X s , then
39
Now, let , therefore
sX s 39Y s X s U s
11 35 250
sX s 108Y s U s
s s
Let X s U s , then 11
Let X s , or 35
Let , then 250 . If , then:
4
39
108
11
35 36
250
103
or
4 1 0 0 0 0 0
39 0 1 0 0 0 1
108 0 0 1 0 0 0
0 0 0 0 1 0 11
0 0 0 0 0 1 35
0 0 0 0 0 0 250
10-3) (a) Alternatively use equations (10-225), (10-232) and (10-233)
3
3 2
Thestatevariablesaredefinedas
x1 ( t ) = y ( t )
dy ( t )
x2 ( t ) =
dt
Thenthestateequationsarerepresentedbythevectormatrixequation
dx ( t )
= Ax ( t ) + Bu ( t )
dt
wherex(t)isthe21statevector,u(t)thescalarinput,and
0 1 3
A= B=
2 3 1 (Alsoseesection233or106)
C = [1 0] D=0
G ( s ) = C ( sI A ) B + D
1
104
MATLAB
>> clear all
>> syms s
>> A=[0,1;-2,-3]
A=
0 1
-2 -3
>> B=[0;1]
B=
>> C=[3,1]
C=
3 1
>> s*eye(2)-A
ans =
[ s, -1]
[ 2, s+3]
>> inv(ans)
ans =
[ (s+3)/(s^2+3*s+2), 1/(s^2+3*s+2)]
[ -2/(s^2+3*s+2), s/(s^2+3*s+2)]
>> C*ans*B
ans =
3/(s^2+3*s+2)+s/(s^2+3*s+2)
105
Use ACSYS as demonstrate in section 10-19-2
1) Activate MATLAB
2) Go to the folder containing ACSYS
3) Type in Acsys
4) Click the Transfer Function Symbolic pushbutton
5) Enter the transfer function
6) Use the State Space option as shown below:
You get the next window. Enter the A,B,C, and D values.
106
-------------------------------------------------------------
State Space Analysis
-------------------------------------------------------------
Inputs:
A=| 0 1| B=|0|
|-2 -3| |1|
C=|3 1| D=|0|
State Space Representation:
Dx = | 0 1|x + |0|u
|-2 -3| |1|
107
y = |3 1|x + |0|u
Determinant of (s*I-A):
s +3s+2
Characteristic Equation of the Transfer Function:
s +3s+2
The eigen values of A and poles of the Transfer Function are:
-1
-2
Inverse of (s*I-A) is:
[ s+3 1 ]
[ ------------ ------------]
[ 2 2 ]
[s +3s+2 s + 3 s + 2]
[ ]
[ 2 s ]
[- ------------ ------------]
[ 2 2 ]
[ s +3s+2 s + 3 s + 2]
State transition matrix (phi) of A:
[ 2 exp(-t) - exp(-2 t) exp(-t) - exp(-2 t) ]
[ ]
[-2 exp(-t) + 2 exp(-2 t) -exp(-t) + 2 exp(-2 t)]
Transfer function between u(t)and y(t) is:

108
s+3
------------
s +3s+2
No Initial Conditions Specified
States (X) in Laplace Domain:
[ 1 ]
[---------------]
[(s + 2) (s + 1)]
[ ]
[ s ]
[---------------]
[(s + 2) (s + 1)]
Inverse Laplace x(t):
[ exp(-t) - exp(-2 t) ]
[ ]
[-exp(-t) + 2 exp(-2 t)]
Output Y(s):
s+3
---------------
(s + 2) (s + 1)
Inverse Laplace y(t):
2 exp(-t) - exp(-2 t)
109
Use the same procedure for parts b, c and d.
10-4) a) x 2 , 2

.
0.5 0.5
As a result:
2 0 0 1 1
0.5 0 0 0 0
0 1 0 0 0
1 0 1 0 0
1 1 1 0
b) 2 2
4 3 4 3 2 3
As a result:
2 0 0 1
2 3 0 1
1 1 6 0
1 0 1
c) 5 5
2 2
4 4
1010
2 2
As a result:
5 0 0 0 1
1 2 1 0 1 u
0 1 4 0 0
0 0 2 0 0
0 1 0 1
10-5)Weshallfirstshowthat
2
I A 1 A
( s ) = ( sI A ) =
1
+ 2
+ 2
+L
s s 2! s
Wemultiplybothsidesoftheequationby ( sI A ) ,andwegetI=I.TakingtheinverseLaplacetransform
onbothsidesoftheequationgivesthedesiredrelationshipfor ( t ) .
10-6)
(a) USEMATLAB
Amat=[0 1;-2 -1]
[mA,nA]=size(Amat);
rankA=rank(Amat);
disp(' Characteristic Polynomial: ')
chareq=poly(Amat);
[mchareq,nchareq]=size(chareq);
syms s;
poly2sym(chareq,s)
[evecss,eigss]=eig(Amat);
disp(' Eigenvalues of A = Diagonal Canonical Form of A is:');
Abar=eigss,
disp('Eigen Vectors are ')
T=evecss
% state transition matrix
ilaplace(inv([s 0;0 s]-Amat))
Results in MATLAB COMMAND LINE

Amat=
01
1011
21
CharacteristicPolynomial:
ans=
s^2+s+2
EigenvaluesofA=DiagonalCanonicalFormofAis:
Abar=
0.5000+1.3229i0
00.50001.3229i
EigenVectorsare
T=
0.20410.5401i0.2041+0.5401i
0.81650.8165
phi=ilaplace(inv([s0;0s]Amat))
phi=
[1/7*exp(1/2*t)*(7*cos(1/2*7^(1/2)*t)+7^(1/2)*sin(1/2*7^(1/2)*t)),2/7*7^(1/2)*exp(
1/2*t)*sin(1/2*7^(1/2)*t)]
[4/7*7^(1/2)*exp(1/2*t)*sin(1/2*7^(1/2)*t),1/7*exp(1/2*t)*(7*cos(1/2*7^(1/2)*t)
7^(1/2)*sin(1/2*7^(1/2)*t))]
%usevpatoconverttodigitalformat.Usedigit(#)toadjustlevelofprecisionifnecessary.
vpa(phi)
ans=
[.1428571*exp(.5000000*t)*(7.*cos(1.322876*t)+2.645751*sin(1.322876*t)),
.7559289*exp(.5000000*t)*sin(1.322876*t)]
[1.511858*exp(.5000000*t)*sin(1.322876*t),
1012
.1428571*exp(.5000000*t)*(7.*cos(1.322876*t)2.645751*sin(1.322876*t))]
ANALYTICALSOLUTION:
2
Characteristicequation: ( s ) = sI A = s + s + 2 = 0
Eigenvalues: s = 0.5 + j1323

. , 0.5 j1323
.
Statetransitionmatrix:
cos1.323t + 0.378 sin 1.323t 0.756 sin 1.323t 0.5 t

(t ) =
1.069 sin (1.323t 69.3 )
e
1.512 sin 1.323t
o

Alternatively
USE ACSYS as illustrated in section 10-19-1

1) Activate MATLAB
3) Type in Acsys
4) Click the State Space pushbutton
5) Enter the A,B,C, and D values. Note C must be entered here and must have the same number of
columns as A. We us [1,1] arbitrarily as it will not affect the eigenvalues.
6) Use the Calculate/Display menu and find the eigenvalues.
1013
FromMATLABCommandWindow:
TheAmatrixis:
Amat=
01
21
1014
ans=
s^2+s+2
Abar=
0.5000+1.3229i0
00.50001.3229i
EigenVectorsare
T=
0.20410.5401i0.2041+0.5401i
0.81650.8165
THERESTARESAMEASPARTA.
2
(b)Characteristicequation: ( s ) = sI A = s + 5s + 4 = 0 Eigenvalues: s = 4, 1
1.333e t 0.333e 4 t t
0.333e 0.333e 4 t
(t ) = t 4 t t 4 t
1.333e 1.333e 0.333e + 1.333e
(c)Characteristicequation: ( s ) = ( s + 3 ) = 0 Eigenvalues: s = 3, 3
2
e 3t 0
(t ) = 3 t
0 e
Eigenvalues: s = 3, 3
2
(d)Characteristicequation: ( s ) = s 9 = 0
e3t 0
(t ) = 3 t
0 e
1015
Eigenvalues: s = j 2, j 2
2
(e)Characteristicequation: ( s ) = s + 4 = 0
cos 2t sin 2t
(t ) =
sin 2t cos 2t
Eigenvalues: s = 1, 2, 2
3 2
(f)Characteristicequation: ( s ) = s + 5 s + 8 s + 4 = 0
e t 0 0

(t ) = 0
2 t 2 t

e te

0 0 e
2 t
(g)Characteristicequation: ( s ) = s + 15 s + 75 s + 125 = 0 Eigenvalues: s = 5, 5, 5

3 2
e 5 t te
5 t
0

(t ) = 0
5 t 5 t

e te

0 0 e
5 t
Statetransitionequation: x(t ) = (t ) x(t ) + (t )Br ( ) d

t
(t ) foreachpartisgiveninProblem53.
0
10-7)In MATLAB USE ilaplace to find L1 ( sI A ) BR ( s ) see previous problem for codes.
1
(a)
1 s + 1 1 0 1 1 1
(t )Br ( )d = L ( sI A ) BR ( s ) = L
t 1
1 1

0
( s ) 2 s 1 0 1 s
s+2
s ( s2 + s + 2)
=
1 + 0.378 sin 1.323t cos1.323t
=L
1
t0
s2 1 + 1.134 sin 1.323t + cos1.323t
s s2 + s + 2
( )
1016
(b)
1 s + 5 1 1 1 1
(t )Br ( )d = L ( sI A ) BR ( s ) = L
t 1
1 1

0
( s ) 4 s 1 1 s
s+6 1.5 1.67 + 0.167
s ( s + 1)( s + 2) 1.5 1.67e t + 0.167e 4 t
=L
1
= L1 s s + 1 s + 4 = t0
s4 1 + 1.67 0.667 1 + 1.67e 0.667e 4t
t
s ( s + 1)( s + 4) s s + 1 s + 4
(c)
1 0

1 s + 3
0 1
0 (t )Br ( )d = L ( sI A ) BR( s) = L
t 1
1

0 1 1 s
s + 3
0
=L
1 = 0
t0

1 3 t
0.333 (1 e )
s ( s + 3 )
(d)
1 0

1 s 3
0 1
0 ( )
t 1
L L
1
= =
1 1 s
( t ) B r ( ) d s I A B R ( s )
0

s + 3
0
=L
1 = 0
t0
0.333 (1 e 3t )

1

s ( s + 3 )
(e)
1 2

2
1 s + 4
0 1
0 ( )
t 1

1
= L = L
s 1 s
( t ) B r ( ) d s I A B R ( s )
2
s 2 + 4
s + 4
2

2
s 2
=L =
1
t 0

1 0.5 sin 2t
( s 2 + 4 )
(f)
1017
1
0 0
s + 1 0
1 1 1 1 1
0 (t )Br ( )d = L ( sI A ) BR( s ) = L 0
t 1
1
2
s+2 ( s + 2) s
0
1
0

0
s+2
0
0
=L
1 1 2 t
= 0.5 (1 e ) t 0
s ( s + 2)
0
0
(g)
1 1
0

s + 5 ( s + 5) 2
0
1 1 1 1
0 (t )Br ( ) d = L ( sI A ) BR ( s ) = L 0
t 1
1
2
0
s+5 ( s + 5) s
1
1
0 0
s+5

0

0
0
1
1 1 0.04 0.04 0.2 5 t 5 t
=L =L
2 s s + 5 ( s + 5 )2 = 0.04 0.04e 0.2te u s (t )
s ( s + 5) 0.2 0.2e
5 t

1 0.2 0.2

s ( s + 5 ) s s+5
10-8)Statetransitionequation: x(t ) = (t ) x(t ) + 0 (t )Br ( ) d (t ) foreachpartisgiveninProblem53.

t
(a)
1018
1 s + 1 1 0 1 1 1
(t )Br ( )d = L ( sI A ) BR ( s ) = L
t 1
1 1

0
( s ) 2 s 1 0 1 s
s+2
s ( s + s + 2)
2
=
1 + 0.378 sin 1.323t cos1.323t
=L
1
t0
s2 1 + 1.134 sin1.323t + cos1.323t

s s2 + s + 2
( )
(b)
1 s + 5 1 1 1 1
(t )Br ( ) d = L ( sI A ) BR ( s ) = L
t 1
1 1

0
( s ) 4 s 1 1 s
s+6 1.5 1.67 + 0.167
s ( s + 1)( s + 2) s s + 1 s + 4 1.5 1.67e t + 0.167e 4 t
=L =L
1 1
= t0
s4 1 + 1.67 0.667 1 + 1.67e 0.667e 4t
t
s ( s + 1)( s + 4) s s + 1 s + 4
(c)
1 0

1 s + 3
0 1
0 ( )
t 1
L L
1
= =
1 1 s
( t ) B r ( ) d s I A B R ( s )
0

s + 3
0
=L
1 = 0
t0

1 3 t
0.333 (1 e )
s ( s + 3 )
(d)
1 0

1 s 3
0 1
0 (t )Br ( )d = L ( sI A ) BR( s ) = L
t 1
1

0 1 1 s

s + 3
0
=L
1 = 0
t0

1 3 t
0.333 (1 e )
s ( s + 3 )
(e)
1019
1 2

2
1 s + 4
0 1
0 ( )
t 1

1
= L = L
s 1 s
( t ) B r ( ) d s I A B R ( s )
2
s 2 + 4
s + 4
2

2
s 2
=L =
1
t 0

1 0.5 sin 2t
( s + 4 )
2
(f)
1
0 0
s + 1 0
1 1 1 1 1
0 (t )Br ( )d = L ( sI A ) BR( s ) = L 0
t 1
1
2
s+2 ( s + 2) s
0
1
0

0
s+2
0
0
=L
1
1
2 t
= 0.5 (1 e ) t 0
s ( s + 2)
0
0
(g)
1 1
0

s + 5 ( s + 5) 2
0
1 1 1 1
0 (t )Br ( ) d = L ( sI A ) BR ( s ) = L 0
t 1
1
2
0
s+5 ( s + 5) s
1
1
0 0
s+5

0
0
0
1
1 1 0.04 0.04 0.2 5 t 5 t
=L =L

( + ) 2 s s + 5 ( s + 5 )2 = 0.04 0.04e 0.2te u s (t )
s s 5
5 t
0.2 0.2e
1 0.2 0.2

s ( s + 5 ) s s+5
1020
10-9)(a)Notastatetransitionmatrix,since ( 0) I (identitymatrix).
(b)Notastatetransitionmatrix,since ( 0) I (identitymatrix).
(c) ( t ) isastatetransitionmatrix,since ( 0 ) = I and
1
1
0 1 0
[ (t ) ] = t
1
t
= = ( t )
1 e e 1 e
t t
e
(d) ( t ) isastatetransitionmatrix,since ( 0) = I ,and
e2t te
2t
t e / 2
2 2t
[ (t ) ]1 = 0 e
2t
te
2t
= ( t )
0 0 e
2t

1 3 1
10-10) a) and
2 3 2
Therefore:
2
2 2 2
If 0 0, then 0.5 0.5
b)
0.5
. 1 1
. .
0.5 0.5 0.5
. .
2 0.5 5 0.5
If x 0 0, then
.
0 1 0.5 0.5 0.5 0.5
2 2 .
0.5
.
5
.
0.5 0.5 1
and
x .
y t 1 0 x x e sin 0.5t
10-11)(a)(1)EigenvaluesofA:2.325, 0.3376 + j 0.5623, 0.3376 j 0.5623

1021
(2)Transferfunctionrelation:
s 2 + 3s + 2 1 0
1
s 1 0 0 s+3 1
X( s ) = ( sI A ) BU ( s ) =
1 1
0 s 1 0 U ( s ) = 1 1
s ( s + 3) s 0 U (s) =
1
s U (s)
( s) (s) (s) 2
1 2 s + 3 1 s 2 s 1 s 1
2
s
3 2
( s ) = s + 3 s + 2 s + 1
(3)Outputtransferfunction:
1
Y ( s) 1 1
= C( s ) ( sI A ) B = [1 0]
1
0 s = 3
( s ) 2 s + 3s + 2 s + 1
2
U (s)
s

7) Activate MATLAB
9) Type in Acsys
12) Use the Calculate/Display menu and find the eigenvalues and other State space calculations.
1022
TheAmatrixis:
Amat=
010
001
123
1023
ans=
s^3+3*s^2+2*s+1
Abar=
2.324700
00.3376+0.5623i0
000.33760.5623i
EigenVectorsare
T=
0.16760.78680.7868
0.38960.2657+0.4424i0.26570.4424i
0.90560.15910.2988i0.1591+0.2988i
StateSpaceModelis:
a=
x1x2x3
x1010
x2001
x3123
b=
u1
x10
x20
x31
1024
c=
x1x2x3
y1100
d=
u1
y10
Continuoustimemodel.
ans=
s^3+3*s^2+2*s+1
EquivalentTransferFunctionModelis:
Transferfunction:
1.776e015s^2+6.661e016s+1
s^3+3s^2+2s+1
Pole,ZeroForm:
Zero/pole/gain:
1.7764e015(s^2+0.375s+5.629e014)
(s+2.325)(s^2+0.6753s+0.4302)
Thenumeratorisbasicallyequalto1
1025
Usethesameprocedureforotherparts.
(b)(1)EigenvaluesofA: 1, 1.
1
1 s + 1 1 0 ( s + 1)2
X ( s ) = ( sI A ) BU ( s ) =
1
= U (s) ( s) = s + 2s + 1
2
( s ) 0 s + 1 1
U ( s )
1
( s + 1)

1
Y ( s) ( s + 1)2 1 1 s+2
= C( s ) ( sI A ) B = [1 1] =
1
+ =
1 ( s + 1) s + 1 ( s + 1)
2 2
U (s)
s + 1
(c)(1)EigenvaluesofA: 0, 1, 1.
s 2 + 2s = 1 s+2 1 0 1
1
( )
1
X ( s ) = ( s I A ) BU ( s ) = s ( s + 2) ) s 0 U ( s ) =
1
(s ) = s s + 2s + 1
2
0 s U (s)
( s) ( s) 2
0 s s 1
2
s
1
s +1
= C( s ) ( sI A ) B = [1 1 0 ] s =
Y (s) 1 1
=
U (s) 2 s ( s + 1) s ( s + 1)
2
2
dy dx1 dx 2 d y dx 2 dx3
10-12) Wewrite = + = x 2 + x3 = + = x1 2 x 2 2 x3 + u
dt dt dt dt dt dt
1026
dx1
dt
0 1 0 x1 0
d x dy
= = 0 1 1 x2 + 0 u (1)
dt dt
2 1 2 2 x3 1
d y
dt 2
x1 1 0 0 1 0 0
x = y = 1 1 0 x x = 1 1 0 x (2)

y& 0 1 1 1 1 1
SubstituteEq.(2)intoEq.(1),wehave
1 1 0 0
= A 1 x + B 1u = 0 0 1 x = 0 u

dx

dt
1 0 2 1
10-13)ForMATLABCodessee1015
(a)
s 2 0
3 2 3 2
sI A = 1 s 2 0 = s 3 s + 2 = s + a2 s + a1 s + a0 a0 = 2, a1 = 0, a2 = 3
1 0 s 1
a1 a2 1 0 3 1 0 2 4
M = a2 1 0 = 3 1 0
S = B AB A B = 1 2 6
2

1 0 0 1 0 0 1 1 1
2 2 0
P = SM = 0 1 1

4 2 1
1027
(b)
s 2 0
3 2 3 2
sI A = 1 s 1 0 = s 3s + 2 = s + a2 s + a1 s + a0 a0 = 2, a1 = 0, a2 = 3
1 1 s 1
a1 a2 1 0 3 1 1 2 6
M = a2 1 0 = 3 1 0
S = B AB A B = 1 3 8
2

1 0 0 1 0 0 0 0 1
0 1 1
P = SM = 1 0 1

1 0 0
(c)
s+2 1 0
3 2 3 2
sI A = 0 s+2 0 = s + 7 s + 16 s + 12 = s + a2 s + a1 s + a0 a0 = 12, a1 = 16, a2 = 7
1 2 s+3
a1 a2 1 16 7 1 1 1 0
M = a2 0 = 7 1 0
S = B AB A B = 1 2 4
2
1

1 0 0 1 0 0 1 6 23
9 6 1
P = SM = 6 5 1

3 1 1
(d)
s +1 1 0
3 2 3 2
sI A = 0 s =1 1 = s + 3 s + 3s + 1 = s + a2 s + a1 s + a0 a0 = 1, a1 = 3, a2 = 3
0 0 s +1
1028
a1 a2 1 3 3 1 0 1 1
M = a2 1 0 = 3 1 0 S = B AB A B = 1 0 1
2

1 0 0 1 0 0 1 1 1
2 1 0
P = SM = 2 3 1

1 2 1
(e)
s 1 1
2 2
sI A = = s + 2 s 1 = s + a1 s + a0 a0 = 1, a1 = 2
2 s+3
a1 1 2 1 0 1
M= = S = [B AB ] =
1 0 1 0

1 3
1 0
P = SM =
1 1
10-14)ForMATLABcodessee1015
(a)FromProblem1013(a),
0 3 1
M = 3 1 0

1 0 0
Then,
C 1 0 1 0.5 1 3
V = CA = 1 2 1 Q = ( MV ) = 0.5 1.5 4
1

2
CA 1 2 1 0.5 1 2
(b)FromProblem1013(b),
16 7 1
M = 7 1 0

1 0 0
1029
C 1 0 1 0.2308 0.3077 1.0769

V = CA = 1 3 1 Q = ( MV ) = 0.1538 1.3846
1
0.5385
2
CA 2 5 1 0.2308 0.3077 0.0769
(c)FromProblem1013(c),
C 1 0 0
V = CA = 2 1 0
2
CA 4 4 0
SinceVissingular,theOCFtransformationcannotbeconducted.
(d)FromProblem1013(d),
3 3 1
M = 3 1 0

1 0 0
Then,
C 1 0 1 1 1 0
V = CA = 1 1 1 Q = ( MV ) = 0 1 2
1

2
CA 1 2 2 1 1 1
(e)FromProblem1013(e),
2 1 C 1 0 0 1
Q = ( MV ) =
1
M= Then, V = 2
=
1 0 CA 1 1 1 3
10-15)(a)EigenvaluesofA: 1,2.7321,0.7321
0 0.5591 0.8255
T = [ p1 p2 p 3 ] = 0 0.7637 0.3022

1 0.3228 0.4766
1030
where p1 , p 2 , and p 3 aretheeigenvectors.
(b)EigenvaluesofA: 1,2.7321,0.7321
0 0.5861 0.7546
T = [ p1 p2 p 3 ] = 0 0.8007 0.2762

1 0.1239 0.5952
where p1 , p 2 , and p 3 aretheeigenvalues.
(c)EigenvaluesofA: 3,2,2.AnonsingularDFtransformationmatrixTcannotbefound.
(d)EigenvaluesofA: 1,1,1
ThematrixAisalreadyinJordancanonicalform.Thus,theDFtransformationmatrixTistheidentity
matrixI.
(e)EigenvaluesofA: 0.4142,2.4142
0.8629 0.2811
T = [p 2 p2 ] =
0.5054 0.9597

1) Activate MATLAB
3) Type in Acsys
7) Next use the Calculate/Display menu and conduct State space calculations.
8) Next use the Calculate/Display menu and conduct Controlability calculations.
NOTE: the above order of calculations MUST BE followed in the order stated, otherwise you will get an
error.
1031
SOLVE PART (a)
The A matrix is:
1032
Amat =
0 2 0
1 2 0
-1 0 1
Characteristic Polynomial:
ans =
s^3-3*s^2+2
Eigenvalues of A = Diagonal Canonical Form of A is:
Abar =
1.0000 0 0
0 2.7321 0
0 0 -0.7321
1033
Eigen Vectors are
T=
0 0.5591 0.8255
0 0.7637 -0.3022
1.0000 -0.3228 0.4766
State-Space Model is:
a=
x1 x2 x3
x1 0 2 0
x2 1 2 0
x3 -1 0 1
b=
u1
x1 0
x2 1
x3 1
c=
x1 x2 x3
y1 0 1 1
1034
d=
u1
y1 0
Continuous-time model.
ans =
s^3-3*s^2+2
Equivalent Transfer Function Model is:
Transfer function:
2 s^2 - 3 s - 4
------------------------------
s^3 - 3 s^2 + 8.882e-016 s + 2
Pole, Zero Form:
Zero/pole/gain:
2 (s-2.351) (s+0.8508)
--------------------------
1035
(s-2.732) (s-1) (s+0.7321)
The Controllability Matrix [B AB A^2B ...] is =
Smat =
0 2 4
1 2 6
1 1 -1
The system is therefore Controllable, rank of S Matrix is =
rankS =
Mmat =
0 -3 1
-3 1 0
1 0 0
The Controllability Canonical Form (CCF) Transformation matrix is:
1036
Ptran =
-2 2 0
0 -1 1
-4 -2 1
The transformed matrices using CCF are:
Abar =
0 1.0000 0.0000
0 -0.0000 1.0000
-2.0000 0.0000 3.0000
Bbar =
Cbar =
-4 -3 2
Dbar =
0
1037
10-16) a)
Consider:
2 2
Therefore:
0 1 0 0
0 0 1 0 0
0
0 0 0 1
1
0 0 2 0
1 0 1 0
U U
As , therefore sY s Y s
U U
Let X s Y s . If y x , then sY s sX s X , or x x . As a
result:
Now consider X , and , then
Therefore:
0 1 0 0 0
2 0 1 0 1 u
0 0 0 1 0
0 0 0 0 1
b)
Consider:
2
1038
4 4
Therefore:
0 1 0
4 4 1
1 2
As , therefore 4 2 . As a result:
4 2
21
4 4
10-17)ForMATLABcodessee1015
(a)
1 2
S = [B AB ] = S is singular.
0 0
(b)
1 1 1
S = B AB A B = 2 2 2 S is singular.
2

3 3 3
(c)
2 2 + 2 2
S = [B AB ] = S is singular.
2 2+ 2
(d)
1 2 4
S = B AB A B = 0 0 0 S is singular.
2

1 4 14
10-18) a, d and e are controllable
b, c, and f are not controllable
1039

9) Activate MATLAB
11) Type in Acsys
16) Next use the Calculate/Display menu and conduct Controlability calculations.
NOTE: the above order of calculations MUST BE followed in the order stated, otherwise you will get an
error.
1040
For part b, the system is not Controllable because [B AB] is singular (rank is less than 2):
The A matrix is:
Amat =
-1 0
0 -2
ans =
s^2+3*s+2
Abar =
-2 0
0 -1
Eigen Vectors are
T=
0 1
1 0
ans =
s^2+3*s+2
Transfer function:
-----
s+1
Pole, Zero Form:

1041
Zero/pole/gain:
-----
(s+1)
Smat =
2 -2
0 0 Rank is 1, and this is a singular matrix
The system is therefore Not Controllable, rank of S Matrix is =
rankS =
Mmat =
3 1
1 0
Ptran =
4 2
0 0
10-19) a, d, and e are observable
b, c, and f are not observable
Using ACSYS (also see the previous problem for more details):
1042
For part b, the system is not observable. Note: you must choose a B matrix arbitrarily.
The A matrix is:
Amat =
-1 0
0 -2
ans =
s^2+3*s+2
1043
Abar =
-2 0
0 -1
Eigen Vectors are
T=
0 1
1 0
ans =
s^2+3*s+2
Transfer function:
Pole, Zero Form:
Zero/pole/gain:
The Observability Matrix (transpose:[C CA CA^2 ...]) is =
Vmat =
0 1
0 -2
The System is therefore Not Observable, rank of V Matrix is =
rankV =
1
1044
Mmat =
3 1
1 0
10-20)(a)Rewritethedifferentialequationsas:
d m B d m K b d m
2 2
K Ki dia Ra Ka Ks
2
= 2
m + ia = ia + ( r
m )
dt J dt J J dt La dt La La
d m
Statevariables: x1 = m , x 2 = , x3 = ia
dt
dx1

dt 0 0 0
1
x1
K Ki
2 =
dx B
x2 + 0 r y= 1 0 0 x = x1
dt J J
x K K
J
dx K K Ra 3 a s
3 a s
Kb
La
dt La La La
(b)Forwardpathtransferfunction:
1

s
1 0 0

m (s) K B Ki Ki K a
G ( s ) = = [1 0 0 ] s+ 0 =
J J
E (s)

J
K o (s)
0 a
s+ a
Kb R
La La La
o ( s ) = JLa s + ( BLa + Ra J ) s + ( KLa + K i K b + Ra B ) s + KRa = 0

3 2

Closedlooptransferfunction:
1045
1

s
1 0 0

m (s) K B Ki K sG(s)
M (s) = = [1 0 0 ] s+ 0 =
r (s) J J

J
K K 1 + K s (s)
Ka Ks a s
s+ a
Kb R
La
La La La
Ki K a K s
=
JLa s + ( BLa + Ra J ) s + ( KLa + K i K b + Ra B ) s + K i K a K s + KRa
3 2
10-21)(a)
0 1 1 0 0 1 1 0
A= A = A = A =
2 3 4
1 0 0 1 1 0 0 1
(1)Infiniteseriesexpansion:
t2 t4 t
3
t
5

1 + L t + L
1 2 2 2! 4! 3! 5! cos t sin t
(t ) = I + At + A t + L = =
sin t cos t
3 5 2 4
2! t t t t
t + 3! 5! + L 1 2! + 4! L
(2)InverseLaplacetransform:
1
s 1 1 s 1 cos t sin t
( s ) = ( sI A ) =
1
= 2 (t ) =
1 s s + 1 1 s sin t cos t
(b)
1 0 1 0 1 0 1 0
A= A = A = A =
2 3 4
0 2 0 4 0 8 0 16
t
2
t
3
t
4

1 t + + +L 0 e t
1 2 2 2! 3! 4! 0
(t ) = I + At + A t + L = = 2 t
0 e
2 3
2! 4t 8t
1 2t + +L

0
2! 3!
1046
1 0

s + 1 0 s +1 1
et 0
( s ) = ( sI A ) =
1
= (t ) =
0 s + 2
2 t
0 1 0 e
s + 2
(c)
0 1 1 0 0 1 1 0
A= A = A = A =
2 3 4
1 0 0 1 1 0 0 1
t2 t4 t
3
t
5

1 + + + L t + + L
1 2 2 2! 4! 3! 5! e t + et e + e
t t
(t ) = I + At + A t + L = = t t
e + e e +e
3 5 2 4 t t
2! t t t t
t + + L 1+ + + L
3! 5! 2! 4!
0.5 0.5 0.5

+

0.5
s 1
1
1 s +1 s 1
s 1 s + 1 s 1
( s ) = ( sI A ) =
1
1 = =
s s 1 1 s 0.5 0.5
2
0.5 0.5
+ +
s + 1 s 1 s + 1 s 1
e t + et e + e
t t
(t ) = 0.5 t t
e + e e +e
t t
10-22)(a) e = K s ( r y ) ea = e e s e s = R s ia eu = Kea
eu eb d y d y
2
ia = eb = K b Tm = K i ia = ( J m + J L )
Ra + Rs
2
dt dt
d y
Solvefor ia intermsof y and , wehave
dt
d y
KK s ( r y ) K b
ia = dt
Rs + Rs + KRs
1047
Differentialequation:
d y d y
2
K i ia Ki
= = Kb KK s y + KK s y
dt
2
Jm + JL ( J m + J L )( Ra + Rs + KRs ) dt
d y
Statevariables: x1 = y , x 2 =
dt
Stateequations:
dx1 0 1 0
dt x1
= KK s K i Kb Ki + KK s K i
dx2 ( J + J )( R + R + KR ) x2 r
( J m + J L )( Ra + Rs + KRs ) ( J m + J L )( Ra + Rs + KRs )
m L
dt
a s s
0 1 x1 0
= + r
322.58 80.65 x2 322.58
Wecanlet v ( t ) = 322.58 r , thenthestateequationsareintheformofCCF.
(b)
1
s 1 1 s + 80.65 1
( sI A ) = 1
= 2
322.58 s + 80.65 s + 80.65 s + 322.58 322.58 s
0.06 1.059 0.014

+
0.014
s + 76.42 s + 4.22 s + 76.42 s + 4.22
=
4.468 4.468 1.0622

0.0587
s + 76.42 s + 4.22 s + 76.42 s + 4.22
Foraunitstepfunctioninput, u s ( t ) = 1 / s.
322.2 1 0.0584 1.058

+
+ +
= s s + 76.42 s + 4.22
1 s ( s 76.42)( s 4.22)
( sI A )1 B =
s 322.2 4.479 + 4.479
s ( s + 76.42)( s + 4.22) s + 76.42 s + 4.22
0.06e 76.42 t 1.059e 4.22 t 0.014e

76.42 t
+ 0.01e
4.22 t

x (t ) = 76.42 t 4.22 t 76.42 t 4.22 t x(0)
4.468e 4.468e 1.0622e 0.0587 e
1 + 0.0584e 76.42 t 1.058e 4.22 t

= 76.42 t 4.22 t t 0
4.479e + 4.479e
1048
2
(c)Characteristicequation: ( s ) = s + 80.65 s + 322.58 = 0
(d) Fromthestateequationsweseethatwheneverthereis Ra there is (1 + K ) Rs . Thus,thepurposeof Rs is

toincreasetheeffectivevalueof Ra by (1 + K ) Rs . Thisimprovesthetimeconstantofthesystem.
10-23)(a)Stateequations:
dx1 0 1 0
dt x1
= KK s K i Kb Ki +
x2
KK s K i
dx2 J ( R + R + KR ) J ( R + Rs + KRs )
r
J ( R + Rs + KRs )

dt
s s
0 1 x1 0
= + r
818.18 90.91 x2 818.18
Let v = 818.18 r . TheequationsareintheformofCCFwithvastheinput.
1
s 1 1 s + 90.91 1
( sI A ) =
1
(b) =
( s + 10.128 )( s + 80.782 ) 818.18 s

818.18 s + 90.91
1.143e 10.128 t 0.142e 80.78 t 0.01415e

10.128 t
0.0141e
80.78 t
x1 (0)
x (t ) = x (0)
2
10.128 t 80.78 t 10.128 t 80.78 t
11.58e + 0.1433e 0.1433e + 1.143e
11.58e 10.128 t
11.58e
80.78 t
+ 10.128 t 80.78 t
t 0
1 1.1434e + 0.1433e
2
(c)Characteristicequation: ( s ) = s + 90.91s + 818.18 = 0
Eigenvalues: 10.128,80.782
(d)Sameremarkasinpart(d)ofProblem514.
10-24) If and P diagonalizing A, let consider , therefore or
The solution for is 0 , therefore
0 1
on the other hand
1049
A Chaptter10Solutio
ons Golnarraghi,Kuo
0 2
From
m equation (1
1) and (2):
1
10-25) Consider and . If , then or
The solution
s for x is 0 , theerefore:
0 (1))
On thhe other hand

d:
0 2
From
m equation (1
1) and (2):
1
1026(a)Fo
orwardpathtrransferfunctio
on: C
Closedlooptr
ransferfunctio
on:
Y (s) 5 ( K1 + K 2 s ) Y (s) G (s) 5 ( K1 + K 2 s )

G ( s ) = = M (s) = = =
s [ s ( s + 4)( s + 5)) + 10 ] 1 + G(s) s + 9 s + 20 s + (10 + 5 K 2 ) s + 5 K1
4 3 2
E (s) R(s)
(b)Stattediagrambydirectdecomp
position:
Stateequations: Outpu
utequation:
x&1 0 1 0 0 x1
0
x& 0 0 1 0 x 0
2 = 2+ r y = 5 K1 5K2 0 x
x&3 0 0 0 1 x3 0
& K (10

1 + 5 K 2 ) 220 9 x4 1
x4 5
1050
1
(c)Finalvalue: r ( t ) = u s ( t ), R( s ) = .
s
5 ( K1 + K 2 s )
lim y (t ) = lim sY ( s ) = lim =1
s + 9 s + 20 s + (10 + 5 K 2 ) s + 5 K1
4 3 2
t s0 s 0
1027InCCFform,
0 1 0 0 L 0 0
0 0 1 0 L 0 0

A= M M M M O M B = M

L
0 0 0 0 1
0
a0 a1 a2 a3 L an 1 1
s 1 0 0 L 0
0 s 1 0 L 0

sI A = M M M M O M

1
0 0 0 0 0

a0 a1 a2 a3 L s + an
n n 1 n 2
sI A = s + an1 s + an 2 s + L + a1 s + a0
SinceBhasonlyonenonzeroelementwhichisinthelastrow,onlythelastcolumnof adj ( sI A ) is
goingtocontributeto adj ( sI A ) B .Thelastcolumnof adj ( sI A ) isobtainedfromthecofactorsof
'
thelastrowof ( sI A ) .Thus,thelastcolumnof adj ( sI A ) B is 1
2 n 1
s s L s .
2
dy d y
1028(a)Statevariables: x1 = y, x 2 = , x3 = 2

dt dt
Stateequations: x& ( t ) = Ax ( t ) + Br ( t )
1051
0 1 0 0
A=0 0 1 B = 0

1 3 3 1
(b)Statetransitionmatrix:
s 2 + 3s + 3 s + 3 1
1
s 1 0
1
( s ) = ( sI A ) = 0 s 1 =
1
1 s + 3s s
2

(s)
1 3 s + 3 s 3s 1 s
2
1 1 1 1 2 1
+ + +
s + 1 ( s + 1) ( s + 1) ( s + 1) ( s + 1) ( s + 1)
2 3 2 3 3
1 1 1 2 s
= +
( s + 1)3 s +1 ( s + 1) 2
( s + 1) 3
( s + 1)3

s 3
2
+
2 s

( s + 1)3 ( s + 1) 2
( s + 1) 3
( s + 1)3
( s ) = s + 3s + 3s + 1 = ( s + 1)
3 2 3

(1 + t + t 2 / 2 ) e t (t + t ) e 2 t 2
t e /2
t

(t ) =
2 t
t e / 2 (1 + t t ) e 2 t
( t t / 2 ) e
2 t
( t + t 2 / 2 ) e t 2
t e
t
(1 2t + t 2 / 2 ) e t
(c)UseACSYSorMATLABandfollowtheprocedureshowninsolutionto103.
1052
StateSpaceAnalysis
Inputs:
A=|010|B=|0|
|001||0|
|133||1|
C=|100|D=|0|
StateSpaceRepresentation:
|010||0|
Dx=|001|x+|0|u
|133||1|
y=|100|x+|0|u
Determinantof(s*IA):
32
s+3s+3s+1
CharacteristicEquationoftheTransferFunction:
32
1053
s+3s+3s+1
TheeigenvaluesofAandpolesoftheTransferFunctionare:
1
1
1
Inverseof(s*IA)is:
[2]
[s+3s+3s+31]
[]
[%1%1%1]
[]
[1s(s+3)s]
[]
[%1%1%1]
[]
[2]
[s3s+1s]
[]
[%1%1%1]
32
%1:=s+3s+3s+1
Statetransitionmatrix(phi)ofA:
[222]
[1/2exp(t)(2+2t+t),(t+t)exp(t),1/2texp(t)]
[222
[1/2texp(t),(t1+t)exp(t),1/2exp(t)(2t+t)
]
]
[22
[1/2exp(t)(2t+t),exp(t)(3t+t),
2]
1/2exp(t)(24t+t)]
Transferfunctionbetweenu(t)andy(t)is:
32
s+3s+3s+1
NoInitialConditionsSpecified
States(X)inLaplaceDomain:
1054
[1]
[]
[3]
[(s+1)]
[]
[s]
[]
[3]
[(s+1)]
[]
[2]
[s]
[]
[3]
[(s+1)]
InverseLaplacex(t):
[2]
[1/2texp(t)]
[]
[2]
[1/2exp(t)(2t+t)]
[]
[2]
[1/2exp(t)(24t+t)]
OutputY(s):
3
(s+1)
InverseLaplacey(t):
2
1/2texp(t)
StateSpaceAnalysis
Inputs:
A=|010|B=|0|
|001||0|
|133||1|
C=|100|D=|0|
StateSpaceRepresentation:
1055
|010||0|
Dx=|001|x+|0|u
|133||1|
y=|100|x+|0|u
Determinantof(s*IA):
32
s+3s+3s+1
CharacteristicEquationoftheTransferFunction:
32
s+3s+3s+1
TheeigenvaluesofAandpolesoftheTransferFunctionare:
1
1
1
Inverseof(s*IA)is:
[2]
[s+3s+3s+31]
[]
[%1%1%1]
[]
[1s(s+3)s]
[]
[%1%1%1]
[]
[2]
[s3s+1s]
[]
[%1%1%1]
32
%1:=s+3s+3s+1
Statetransitionmatrix(phi)ofA:
[222]
[1/2exp(t)(2+2t+t),(t+t)exp(t),1/2texp(t)]
[222
[1/2texp(t),(t1+t)exp(t),1/2exp(t)(2t+t)
]
]
1056
[22
[1/2exp(t)(2t+t),exp(t)(3t+t),
2]
1/2exp(t)(24t+t)]
Transferfunctionbetweenu(t)andy(t)is:
32
s+3s+3s+1
InitialConditions:
x(0)=1
0
0
States(X)inLaplaceDomain:
[2]
[s+3s+4]
[]
[3]
[(s+1)]
[]
[s1]
[]
[3]
[(s+1)]
[]
[s(s1)]
[]
[3]
[(s+1)]
InverseLaplacex(t):
[2]
[(t+1+t)exp(t)]
[]
[2]
[(t+t)exp(t)]
[]
[2]
[(3t+1+t)exp(t)]
OutputY(s)(withinitialconditions):
2
s+3s+4
1057
3
(s+1)
InverseLaplacey(t):
2
(t+1+t)exp(t)
3 2
(d)Characteristicequation: ( s ) = s + 3 s + 3 s + 1 = 0
Eigenvalues: 1,1,1
dy
1029(a)Statevariables: x1 = y, x 2 =
dt
Stateequations:
dx1 (t )
dt 0 1 x1 (t ) 0
= + r (t )
dx2 (t ) 1 2 x2 (t ) 1
dt
s+2 1
( s + 1)2
1
( s + 1) (1 + t )e t
s 1
2 t
te
( s ) = ( sI A ) =
1
(t ) =
= 1
+
t t
1 s 2 s te (1 t ) e
s +1 2 2
( ) ( s + 1)
( s ) = ( s + 1) = 0
2
Characteristicequation:
dy
(b)Statevariables: x1 = y, x 2 = y +
dt
Stateequations:
2
dx1 dy dx 2 d y dy dy
= = x 2 y = x 2 x1 = 2
+ = y + r = x2 + r
dt dt dt dt dt dt
dx1
dt 1 2 x1 0
= + r
dx2 0 1 x2 1
dt
1058
1 2
s + 1 2 s + 1 ( s + 1)
1 2
e t
te
t
(s) = = (t ) =
0 s + 1 1
t
0 e
0 s +1

(c)Characteristicequation: ( s ) = ( s + 1) = 0 whichisthesameasinpart(a).
2
1030(a)Statetransitionmatrix:
s 1 s
sI A =
s
( sI A ) 1
=
( s ) s
2
(
( s ) = s 2 + +
2 2
)

cos t sin t
(t ) = L ( sI A ) =
1 1 t
e
sin t cos t
(b)EigenvaluesofA: + j , j
1031(a)
3
Y1 ( s ) s 1
= 1 2 3
= 3 2

U1 ( s ) 1+ s + 2s + 3s s + s + 2s + 3
3
Y2 ( s ) s 1 Y1 ( s )
= 1 2 3
= 3 2
=
U 2 ( s) 1+ s + 2s + 3s s + s + 2s + 3 U1 ( s )
(b)Stateequations[Fig.521(a)]: x& = A1x + B1u1 Outputequation: y1 = C1x
0 1 0 0
A1 = 0 0 1 B1 = 0 C1 = [1 0 0]

3 2 1 1
Stateequations[Fig.521(b)]: x& = A 2 x + B 2 u2 Outputequation: y2 = C2x
0 0 3 1
A 2 = 1 0 2 B 2 = 0 C2 = [ 0 0 1]

0 1 1 0
'
Thus, A 2 = A1
1059
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
1032(a)Statediagram:
(b) Statediagram:
1
1033(a)
3
Y ( s) 10 s X ( s)
G( s) = = 1 2 3
Y ( s ) = 100 X ( s )
U ( s) 1 + 8.5 s + 20.5s + 15s X ( s)
1 2 3
X ( s ) = U ( s ) 8.5 s X ( s ) 20.5 s X ( s ) 15 s X ( s )
Statediagraam:
1060
A Chaptter10Solutio
ons Golnarraghi,Kuo
Statteequation: x& ( t ) = Ax ( t ) + Bu ( t )
0 1 0 0

A= 0 0 1 B = 0 AandBarein
A CCF

15 20.5 8.55 1
(b)
3 4
Y ( s) 10 s + 20 s X ( s)
G ( s ) = = 1 2

U ( s) 1 + 4.5 s + 3.5 s X ( s)
3 4 1 2
Y ( s ) = 10 s X ( s ) + 20 s X ( s ) X ( s ) = 4.5 s X ( s ) 3.5 s X ( s ) + U ( s )
Staatediagram:
Staateequations: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 1 0 0 0
0 0 1 0 0
A= B= AandBarein
nCCF
0 0 0 1 0

0 0 3.5 4.55 1
(c)
2 3
Y ( s) 5( s + 1) 5s + 5s X ( s)
G( s) = = = 1 2

U ( s) s( s + 2 )( s + 10) 1 + 12 s + 20 s X ( s)
2 3 1
2
Y ( s) = 5s X ( s ) + 5s X ( s ) X ( s ) = U ( s ) 12 s X ( s ) 20 s X ( s )
1061
A Chaptter10Solutio
ons Golnarraghi,Kuo
Staateequations: x& ( t ) = Ax ( t ) + Bu ( t )
0 1 0 0

A= 0 0 1 B = 0 AandBareinCCF

0 20 12 1
(d)
4
Y (s) 1 s X (s)
G (s) = = =
U (s) (
s ( s + 5) s + 2s + 2
2
) 1
1 + 7 s + 12 s + 10 s
2 3
X (s)
4 1 2 3
Y ( s) = s X ( s) X ( s ) = U ( s ) 7 s X ( s ) 12 s X ( s ) 10 s
Statediagraam:
Stateequations: x& ( t ) = Ax
A ( t ) + Bu ( t )
1062
A Chaptter10Solutio
ons Golnarraghi,Kuo
0 1 0 0 0
0 0 1 0 0
A= B= AandBarreinCCF
0 0 0 1 0

0 10 12 7 1
1
1034(a)
Y ( s) 10 5.71 6.67 0.952
G( s) = = 3 2
= +
U ( s) s + 8.5 s + 20.5s + 15 s + 15 s+2 s+5
A ( t ) + Bu ( t )
1.5 0 0 5.71
A= 0 2 0 B = 6.67

0 0 5 0.952
wtheinputandtheoutputbranchesareallo
ThematrixBisnotunique.Itdependsonhow ocated.
(b)
Y ( s) 10( s + 2) 4.55 0.49 4 5.71

G( s) = = 2
= + + + 2

U ( s) s ( s = 1)(( s + 3.5) s s + 3.5 s +1 s
Stattediagram:
1063
A Chaptter10Solutio
ons Golnarraghi,Kuo
Stateequation: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 1 0 0 0
0 0 0 0 1
A= B=
0 0 1 0 1

0 0 0 3.5 1
(b)
Y ( s) 5( s + 1) 2.5 0.313 0.563

G( s) = = = +
U ( s) s( s + 2)( s + 10) s s+2 s + 10
Sttateequationss: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 0 0 1
A = 0 2 0 B = 1

0 0 10 1
(d)
Y (s) 1 0.1 0.0118 0.0882 s + 0.235

G ( s) = = =
U ( s) (
s ( s + 5) s + 2s + 2
2
) s s+5 s + 2s + 2
2
1064
A Chaptter10Solutio
ons Golnarraghi,Kuo
Sttateequationss: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 0 0 0 1
0 5 0 0 1
A= B=
0 0 0 1 0

0 0 2 2 1
1
1035(a)
Y ( s) 1
10
G( s) = =
U ( s) ( s + 15)(
. s + 2 )( s + 5)
Stattediagram:
A ( t ) + Bu ( t )
5 1 0 0

A = 0 2 1 B=0

0 0 1.5 10
(b)
1065
A Chaptter10Solutio
ons Golnarraghi,Kuo
G(s) =
Y (s)
=
10( s + 2) 100 s + 2 1
=2

s ( s + 1)( s + 3.5) s s + 1 s + 3.5
2
U (s)
Staatediagram:
A ( t ) + Bu ( t )
0 1 0 0 0
0 0 1 1 0
A= B=
0 0 1 1 0

0 0 0 3.5 10
(c)
G ( s) =
Y (s)
=
59 s + 1) 5 s +1 1
=

U (s) s ( s + 2)( s + 10) s s + 2 s + 10
Staatediagram:
A ( t ) + Bu ( t )
0 1 0 0
A = 0 10 1 B = 0

0 0 2 5
(d)
1066
A Chaptter10Solutio
ons Golnarraghi,Kuo
Y (s) 1
G(s) = =
U (s) (
s ( s + 5) s + 2s + 2
2
)
S
Statediagram:

A ( t ) + Bu ( t )
0 1 0 0 0
0 0 1 0 0
A= B=
0 2 2 1 0

0 0 0 5 1
1
1036(a)
3
Y ( s) 10 100 s X ( s)
G( s) = = = 1 2

E ( s) s( s + 4 )( s + 5) 1+ 9s + 20 s X ( s)
(b)
( Dynamicequations:
x&1 0 1 0 x1 0
x& = 0 0 1 x2 + 0 r y = [10
1 0 0] x
2
x&3 10
1 20 9 x3 1
1067
(c)Statetransitionequation:
X 1 (s) s 1 (1 + 9 s 1 + 20 s 2 ) s 2 (1 + 9 s 1 ) s 3 x1 (0) s 3
X (s) = 1 10 s
3 1 2
s (1 + 9 s ) s x2 (0) +
1 1 2 1
s
2
(s) ( s ) 1 s
X 3 ( s ) 10 s
2
20 s
2
s x3 (0)
1
s
1
s 2
+ 9 s + 20 s+9 1 x1 (0) s

s x2 (0) +
1 1
= 10 s ( s + 9)
c (s)
1
c (s)
10 s 10 ( 2 s + 1) s x3 (0)
2
s

1 2 3 3 2
( s) = 1+ 9s + 20 s + 10 s c ( s ) = s + 9 s + 20 s + 10
1.612 0.946 0.114 0.706 1.117 0.169 0.0935 0.171 0.055

5.895 t
x (t ) = 1.14 0.669 0.081 e 4.056 e
x(0)
0.708 t 2.397 t
+ 1.692 2.678 + 0.551 1.009 0.325 e

0.807 0.474 0.057
4.056 6.420 0.972 3.249 5.947 1.915
0.1 0.161e 0.708 t + 0.0706e 2.397 t 0.00935e 5.895 t

+ 0.114e
0.708 t
0.169e
2.397 t
+ 0.055e
5.895 t
t 0
0.087e 0.708 t + 0.406e 2.397 t 0.325e 5.895 t
(d)Output:
(
y (t ) = 10 x1 (t ) = 10 1.612e
0.708 t
0.706e
2.397 t
+ 0.0935e
5.897 t
) x (0) + 10 ( 0.946e 1.117e + 0.1711e ) x (0)
1
0.708 t 2.397 t 5.895 t
2

(
+ 10 1.141e
0.708 t
0.169e
2.397 t
+ 0.0550e
5.895 t
) x (0) + 1 1.61e0.708t + 0.706e2.397t 0.0935e5.895t
3
t0
1037(a)Closedlooptransferfunction:
Y ( s) 10
= 3 2

R( s ) s + 9 s + 20 s + 10
(b)Statediagram:
1068
A Chaptter10Solutio
ons Golnarraghi,Kuo
(c)Stateequation
S s: (d)Stateetransitioneqquations:
[
[Sameanswers
sasProblem5
526(d)]
x&1 0 1 0 x1 0

x& 2 = 0 0 1 x2 + 0 r (e)Outpput:[SameansswerasProbleem526(e)]

x&3 10 20 9 x3 1
1
1038(a)Stattediagram:
(b)Statteequations:
x&1 2 20 1 0 x1 0 1
x& 0 10 1 0 x2 0 0 u
= +
2
x&3 0.1 0 20 1 x3 0 0 TD

&
x4 0 0 0 5 x4 30 0
(c)Tran
nsferfunctionrrelations:
From
mthesystemb
blockdiagram,
1 1
0.2 s
0.3 30e U (s) 90U ( s )
Y ( s ) = TD ( s ) + TD ( s ) + +
( s) s + 2 ( s + 2)( s + 20) ( s + 2)( s + 5)( s + 20) ( s + 5)( s + 20)
0. 2 s 0. 2 s
0.1e ( s + 2)( s + 20) + 0.1e
(
s) = 1+ =
( s + 2 )( s + 20) ( s + 2)( s + 20)
1069
A Chaptter10Solutio
ons Golnarraghi,Kuo
0. 2 s
( s + 19.7 ) 30e + 90( s + 2 )U ( s )
Y ( s ) = 0. 2 s
TD ( s ) +
0. 2 s
( s + 2)( s + 20) + 0.1e ( s + 5) ( s + 2)( s + 20) + 0.1e
0. 2 s
( s + 200) 330e U ( s)
( s ) = 0. 2 s
TD ( s ) +
0. 2 s
( s + 2 )( s + 20) + 0.1e ( s + 5) ( s + 2 )( s + 20) + 0.1e
11
1
1039(a)Th
hereshouldno
otbeanyincom
mingbranchestoastatevariaablenodeotheerthanthe s branch.Thus,we
ouldcreateanewnodeasshowninthefolllowingstatediiagram.
sho
1
(b)State Noticethattherreisaloopwitthgain1afterallthe s brranchesaredeleted,so =2
eequations:N 2.
dx1 17 1 dx 2 15 1 1
= x1 + x2 = x1 x 2 + r Outputequ
uation: y = 6..5 x1 + 0.5 x 2
dt 2 2 dt 2 2 2
1
1040(a)Trransferfunctio
on: (b)Charracteristicequation:
Ks + 5 s + 1
( s + 1) ( s )
2
Y (s)
= + 111s + 2 = 0
2

R(s) ( s + 1) ( s 2
+ 11s + 2 )
Roo
otsofcharacte
eristicequation: 1,0.185,10..82.TheseareenotfunctionsofK.
(c)Whe
enK=1:
2
Y ( s) s + 5s + 1
= 3 2

R( s ) s + 12 s + 133s + 2
Statediaggram:
1070
A Chaptter10Solutio
ons Golnarraghi,Kuo
(d)WheenK=4:
4 s + 5s + 1 ( s + 1)(4 s + 1) 4s + 1
2
Y (s)
= = =
R( s) ( s + 1) ( s 2
+ 11s + 2 ) ( s + 1) ( s 2
+ 11s + 2 ) s + 11s + 2
2
Statediagram:
(e)
Ks + 5 s + 1
( s + 1) ( s )
2
Y (s)
= + 111s + 2 = 0
2

R(s) ( s + 1) ( s 2
+ 11s + 2 )
M
MATLAB
ssolve(s^2+11*ss+2)
aans=11/2+1
1/2*113^(1/2)
11/21/2*1133^(1/2)
>>>vpa(ans)
aans=
.20
10.8
Y (s) Ks 2 + 5s + 1
=
R ( s ) ( s + 1)(
1 s + 0.2)( s + 10.82)
K = 4, 2.1914, 0.4536
P
Polezerocan
ncelationocccursforthegivenvaluesofK.

1071
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
1041(a)
Y (s) 1 100
Gp (s) = = =
U (s) (1 + 0.5s ) (1 + 0.2 s + 0.02 s 2
) s + 12
1 s + 70 s + 1000
3 2
Stattediagrambyddirectdecompposition:
Stateequations:
x&1 0 1 0 x1 0
x& = 0 0 1 x2 + 0 u
2
x&3 100 70 12 x3 1
(b)Charracteristicequuationofcloseddloopsystem
m: R
Rootsofchara
acteristicequation:
5.88, 3.06 + j 4.965, 3.06 j 4.965

3 2
s + 12 s + 70 s + 200 = 0
1
1042(a)
Y ( s) 1 0.066 s 20( s 15)

1
Gp (s) = =
U ( s) (1 + 0.5s ) (1 + 0.133s + 0.0067s 2
) s + 22 s + 190 s + 300
3 2
Stattediagrambydirectdecompposition:
1072
A Chaptter10Solutio
ons Golnarraghi,Kuo
Stateequations:
x&1 0 1 0 x1 0
x& 2 = 0 0 1 x2 + 0

x&3 300 190
1 22 x3 1
Chaaracteristicequuationofcloseedloopsystem
m: R
Rootsofchara
acteristicequation:
12,5+j5,5j5
3 2
s + 22 s + 170 s + 600 = 0
1
1043(a)Statevariables: x1 = m and x 2 = D
Stateequations:
d m Kb Ki + Kb Ra KD KK
Ki d D KD KD
= + D + e = m D
dt JRa J JRa dt JR JR
(b)Stattediagram:
(c)Openlooptransfe
erfunction:
m (s) Ki ( J R s + K D )
KK
=
JJ R Ra s + ( K b J R K i + K D Ra J R + K D JRa ) s + K D K b K i
2
E (s)
Clossedlooptransferfunction:
1073
m (s) K s KK i ( J R s + K D )
=
r (s) JJ R Ra s + ( K b J R K i + K D Ra J R + K D JRa + K s KK i J R ) s + K D K b K i + K s KK i K D
2
(d)Characteristicequationofclosedloopsystem:
( s ) = JJ R Ra s + ( K D J R K i + K D Ra J R + K D JRa + K s KK i J R ) s + K D K b K i + K s KK i K D = 0
2
2
( s ) = s + 1037 s + 20131.2 = 0
Characteristicequationroots: 19.8, 1017.2
1044(a)Stateequations: x& ( t ) = Ax ( t ) + Br ( t )
b d 2 1 0 0 1
A= = B= S = [B AB ] =
c a 2 1 1 1 1
SinceSisnonsingular,thesystemiscontrollable.
(b)
0 d
S = [B AB ] = Thesystemiscontrollablefor d 0.
1 a
1045(a)
1 1 1
S = B AB A B = 1 1 1
2
Sissingular.Thesystemisuncontrollable.

1 1 1
(b)
1 1 1
S = B AB A B = 1 2 4
2
Sisnonsingular.Thesystemiscontrollable.

1 3 9
1046(a)Stateequations: x& ( t ) = Ax ( t ) + Bu ( t )
1074
2 3 1 1 1
A = B= S = [B AB ] = Sissingular.Thesystemisuncontrollable.
1 0 1 1 1
Outputequation: y= 1 0 x = Cx C= 1 0
1 2
V = C A C =
' ' '
Visnonsingular.Thesystemisobservable.
0 3
(b)Transferfunction:
Y ( s) s+3 1
= 2
=
U ( s) s + 2s 3 s 1
Sincethereispolezerocancellationintheinputoutputtransferfunction,thesystemiseither
uncontrollableorunobservableorboth.Inthiscase,thestatevariablesarealreadydefined,andthe
systemisuncontrollableasfoundoutinpart(a).
1047(a) = 1, 2, or 4 .Thesevaluesofwillcausepolezerocancellationinthetransferfunction.
(b)Thetransferfunctionisexpandedbypartialfractionexpansion,
Y ( s) 1 2 4
= +
R( s ) 3( s + 1) 2( s + 2 ) 6( s + 4 )
Byparalleldecomposition,thestateequationsare: x& ( t ) = Ax ( t ) + Br ( t ) ,outputequation: y ( t ) = Cx( t ).
1 0 0 1
A = 0 2 0 B = 2 D=
1
1 1

3 2 6
0 0 4 4
Thesystemisuncontrollablefor=1,or=2,or=4.
(c)Definethestatevariablessothat
1075
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
3
1 0 0
1
A = 0 2 0 B= D = [ 1 2 4]
2
0 0 4 1

6
TThesystemisu or=1,or==2,or=4.
unobservablefo
1
1048
1 b
S = [B A ]= S = ab 1 b 0
2
ab 1
AB
b
Theboundaryoftheregionofcoontrollabilityissdescribedby ab 1 b 2 = 0.
0
Regionsofcontro
ollability:
1
1049
b1 b1 + b2
S = [B A ]= S = 0 when b1b2 b1b2 b2 = 0, or b2 = 0
2
AB
b2 b2
Theesystemiscom ollablewhen b2 0.
mpletelycontro
d1 d2
V = C A C = V = 0 when d1 0.
' ' '

d2 d1 + d 2
mpletelyobservvablewhen d 2 0.
Theesystemiscom
1076
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
1050(a)Stateequations:
dh 1 K I nN Ko d m d m Ki Kb Ki K a
= (q i
qo ) = m h = m = m +
2
ei J = J m + n J L
dt A A A dt d
dt JRa JRa
d m
Stattevariable: x1 = h, x 2 = m , x3 = = m
dt
Statteequations: x& = Ax + Bei
Ko K I nN
0 1 0.016 0 0 0
A A

A= 0 0 1 = 0 0 1 B= 0 = 0

K i K b 0 0 11..767 K K 8333.33

0 0 i a

JRa JRa
Sttatediagram:
(b)Characteristicequ
uationofA:
Ko K I nN
n
s+ 0
A A
sI A = 1

= ss +
Ko s + Ki Kb
0 s = s ( s + 1)( s + 11.767)
1
A JRa
Ki Kb
0 0 s+
JRa
Eige
envaluesofA: 0,1,11.767.
(c)Conttrollability:
1077
0 0 133.33
S = B AB A B = 0 8333.33 98058 S 0. Thesystemiscontrollable.
2

8333.33 98058 1153848
(d)Observability:
(1) C = 1 0 0 :
1 1 1
V = C ( A ) C = 0 0.016 0.016 Visnonsingular.Thesystemisobservable.
' ' ' ' 2 '
AC

0 0 0.016
(2) C = 0 1 0 :
0 0 0
V = C ( A ) C = 1 0 0 Vissingular.Thesystemisunobservable.
' ' ' ' 2 '
AC

0 1 11.767
(3) C = 0 0 1 :
0 0 0
V = C ( A ) C = 0 0 Vissingular.Thesystemisunobservable.
' ' ' ' 2 '
AC 0

1 11.767 138.46
4 2
1051(a)Characteristicequation: ( s ) = sI A = s 25.92 s = 0
Rootsofcharacteristicequation: 5.0912,5.0912,0,0
(b)Controllability:
0 0.0732 0 1.8973
0.0732 0 1.8973 0
S = B

AB
2
A B

A B =
3

0 0.0976 0 0.1728

0.0976 0 0.1728 0

Sisnonsingular.Thus, A , B iscontrollable.
(c)Observability:
1078

(1) C = 1 0 0 0
1 0 25.92 0
0 1 0 25.92
V = C ( A ) C =
' ' ' ' 2 ' ' 3 '
A C (A ) C
0 0 0 0

0 0 0 0
Sissingular.Thesystemisunobservable.

(2) C = 0 1 0 0
0 25.92 0 671.85
1 0 25.92 0
V = C ( A ) C =
' ' ' ' ' ' 3 '
A C (A ) C
2
0 0 0 0

0 0 0 0
(3) C = 0 0 1 0
0 0 2.36 0
0 0 0 2.36
V = C ( A ) C =
' ' ' ' ' ' 3 '
A C (A ) C
2
1 0 0 0

0 1 0 0
Sisnonsingular.Thesystemisobservable.

(4) C = 0 0 0 1
0 2.36 0 61.17
0 0 2.36 0
V = C
'
A C
' ' '
(A ) C
2 '
( A ) C =
' 3 '

0 0 0 0

1 0 0 0
1079
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
1052Thecontrollabilitymatrixis
0 1 0 16 0 3
384
1 0 16 0 384 0

0 0 0 166 0 512
S= R
RankofSis6.Thesystemisccontrollable.
0 0 16 0 512 0

0 1 0 0 0 0

1 0 0 0 0 0
1
1053(a)Trransferfunctio
on:
v ( s) KI H
=
R(s) Jvs
2
(J G
s + KPs + KI + KN
2
)
Stattediagrambydirectdecomp
position:
Statteequations: x& ( t ) = Ax
A ( t ) + Br ( t )
0 1 0 0
0 0
0 1 0 0

A = 0 0 0 1 B=
0
( KI + KN ) KP
0 0 1
JG J G
(b)Characteristicequ
uation: J v s
2
(J G
s + KP s + KI + KN = 0
2
)
1080
A Chaptter10Solutio
ons Golnarraghi,Kuo
1
1054(a)Stateequations: x& (t ) = Ax
A (t ) + Bu1 (t )
3 1 0 0 1
A= B= S = [B B] = 1 2
2 1 AB
0
Sisnonsinggular. A, B iscontrollablee.
Outputequ
uation: y 2 = Cx C = 1 1
1 3
V = C A C =
' ' '
1 3 Vissingu
ular.Thesystem
misunobservaable.

hfeedback, u2 = kc2 , thesttateequationis: x& (t ) = Ax(t ) + Bu1 (t ) .

(b)With
3 2 k 1 0 1
0
A = 1+ g 1+ k B= S= 1+ k
1
0 2 1 2
Sisnonsingularfo
orallfinitevalu
uesofk.Thesyystemiscontro
ollable.
Stattediagram:
Outtputequation: y2 = Cxx C=
1 1

1 + k 1 + k
1 3 + 2k
1 + K (1 + k )
2
V = D' A D =
' '

1 3 + 2k

1 + k (1 + k )
2
Vissingularforanyk.Theesystemwithfeedbackisun
nobservable.
1081
1055(a)
1 2
S = [B AB ] = Sisnonsingular.Systemiscontrollable.
2 7
1 1
V = C A C =
' ' '
Visnonsingular.Systemisobservable.
1 2
(b) u = k1 k 2 x
0 1 k1 k2 k1 1 k2
A c = A BK = =
1 3 2k1 2k 2 1 2 k1 3 2k 2
1 k1 2 k 2 + 2
S = [B AcB] = S = 11 2 k 2 0
2 7 2 k1 4 k 2
11
Forcontrollabillity, k 2
2
1 1 3k1
V = C A c C =
' ' '

1 2 3k 2
Forobservability, V = 1 + 3k1 3k 2 0
1056
Sameas1021(a)
1057
0 x1 0
1
= + r
322.58 80.65 x2 322.58

1082
0 1
A=
322.58 80.65
x& 0 1 x1 0 0
From1022 1 = + r B=
x&2 322.58 80.65 x2 322.58 322.58
C = [1 0]
D=0
UsethestatespacetoolofACSYS
1083
TheAmatrixis:
Amat=
01.0000
322.580080.6500
ans=
s^2+1613/20*s+16129/50
Abar=
4.22060
076.4294
EigenVectorsare
T=
0.23050.0131
0.97310.9999
1084
StateSpaceModelis:
a=
x1x2
x101
x2322.680.65
b=
u1
x10
x2322.6
c=
x1x2
y110
d=
u1
y10
ans=
s^2+1613/20*s+16129/50
Transferfunction:
322.6
s^2+80.65s+322.6
Pole,ZeroForm:
Zero/pole/gain:
322.58
(s+76.43)(s+4.221)
1085
1058
ClosedloopSystemTransferFunction.
Y (s) 1
=
s + ( 4 + k3 ) s + ( 3 + k 2 + k3 ) s + k1
3 2
R(s)
Forzerosteadystateerrortoastepinput, k1 = 1.Forthecomplexrootstobelocatedat1+jand1j,
2
wedividethecharacteristicpolynomialby s + 2 s + 2 andsolveforzeroremainder.
s + ( 2 + k2 )
s + 2 s + 2 s + ( 4 + k3 ) s + ( 3 + k 2 + k3 ) s + 1
2 3 2

s + + 2s
3 2
2s
(2 + k ) s
3
2
+ (1 + k 2 + k 3 ) s + 1
1086
(2 + k ) s
3
2
+ ( 4 + 2 k3 ) s + 4 + 2 k3

( -3+k 2
k3 ) s 3 2k3
Forzeroremainder, 3 2 k3 = 0 Thus k3 = 15
.
3 + k 2 k3 = 0 Thus k 2 = 15
.
Thethirdrootisat0.5.Notalltherootscanbearbitrarilyassigned,duetotherequirementonthe
steadystateerror.
1059(a)OpenloopTransferFunction.
X 1 (s) k3
G ( s) = =
s s + ( 4 + k 2 ) s + 3 + k1 + k 2
2
E (s)
Sincethesystemistype1,thesteadystateerrorduetoastepinputiszeroforallvaluesof k1 , k 2 , and k3
thatcorrespondtoastablesystem.Thecharacteristicequationoftheclosedloopsystemis
s + ( 4 + k 2 ) s + ( 3 + k1 + k 2 ) s + k3 = 0
3 2

Fortherootstobeat1+j,1j,and10,theequationshouldbe:
3 2
s + 12 s + 22 s + 20 = 0
Equatinglikecoefficientsofthelasttwoequations,wehave
4 + k 2 = 12 Thus k2 = 8
3 + k1 + k 2 = 22 Thus k1 = 11
k3 = 20 Thusk3=20
(b)OpenloopTransferFunction.
1087
Y (s) Gc ( s ) 20 20 ( s + 1)( s + 3 )
= = Thus Gc ( s ) =
E ( s) ( s + 1)( s + 3 ) (
s s + 12 s + 22
2
) (
s s + 12 s + 22
2
)
1060(a)
0 1 0 0 0
25.92 0 0 0 0.0732
A = B =

0 0 0 1 0

2.36 0 0 0 0.0976
Thefeedbackgains,from k1 to k 4 :
2.4071E+03 4.3631E+02 8.4852E+01 1.0182E+02
The A B K matrixoftheclosedloopsystem
0.0000E+00 1.0000E+00 0.0000E+00 0.0000E+00
1.5028E+023.1938E+016.2112E+007.4534E+00
0.0000E+00 0.0000E+00 0.0000E+00 1.0000E+00
2/3258E+02 4.2584E+01 8.2816E+00 9.9379E+00
TheBvector
0.0000E+00
7.3200E02
0.0000E+00
1088
9.7600E02
TimeResponses:
1060(b) Thefeedbackgains,from k1 to k2 :
9.9238E+03 1.6872E+03 1.3576E+03 8.1458E+02
0.0000E+00 1.0000E+00 0.0000E+00 0.0000E+00
7.0051E+021.2350E+029.9379E+015.9627E+01
0.0000E+00 0.0000E+00 0.0000E+00 1.0000E+00
9.6621E+02 1.6467E+02 1.3251E+02 7.9503E+01

1089
TheBvector
0.0000E+00
7.3200E02
0.0000E+00
9.7600E02
TimeResponses:
1061ThesolutionsusingMATLAB
1090
(a) Thefeedbackgains,from k1 to k2 :
6.4840E+01 5.6067E+00 2.0341E+01 2.2708E+00
0.0000E+00 1.0000E+00 0.0000E+00 0.0000E+00
3.0950E+023.6774E+01 1.1463E+02 1.4874E+01
0.0000E+00 0.0000E+00 0.0000E+00 1.0000E+00
4.6190E+023.6724E+01 1.7043E+02 1.477eE+01
TheBvector
0.0000E+00
6.5500E+00
0.0000E+00
6.5500E+00
'
(b)TimeResponses: x(0) = 01
. 0 0 0
1091

'
. 0 0 0 , theinitialpositionof x1 or y1ispreturbeddownward
Withtheinitialstates x(0) = 01
fromitsstableequilibriumposition.Thesteelballisinitiallypulledtowardthemagnet,so x3 = y 2 is
negativeatfirst.Finally,thefeedbackcontrolpullsbothbodiesbacktotheequilibriumposition.
'
. 0 , theinitialpositionof x3 or y 2 ispreturbed
Withtheinitialstates x(0) = 0 0 01
downwardfromitsstableequilibrium.Fort>0,theballisgoingtobeattractedupbythemagnet
towardtheequilibriumposition.Themagnetwillinitiallybeattractedtowardthefixedironplate,and
thensettlestothestableequilibriumposition.Sincethesteelballhasasmallmass,itwillmovemore
1092
actively.
1062(a)BlockDiagramofSystem.
u = k1 x1 + k 2 ( x1 + w1 ) dt
StateEquationsofClosedloopSystem:
dx1
dt 2 k1 1 x1 0 1 w1
= +
dx2 k 2 0 x2 k 2 0 w2
dt
1093
s + 2 + k1 1
sI A = = s + ( 2 + k1 ) s + k 2 = 0
2

k2 s
Fors=10,10, sI A = s 2 + 20 s + 200 = 0 Thus k1 = 18 and k 2 = 200
200W1 ( s ) s 2 + s 1W2 ( s ) 200W1 ( s) + sW2 ( s)

X ( s) = X 1 ( s) = 1 1 2
=
1 + 2s + 18 s + 200 s s 2 + 20 s + 200
1 W2
W1 ( s ) = W2 ( s) = W2 = constant
s s
200 + W2 s
X ( s ) = lim x (t ) = lim sX ( s ) = 1
(
s s + 20 s + 200
2
) t s 0
1062(b)WithPIController:
BlockDiagramofSystem:
(K s + K I ) W1 ( s ) + sW2 ( s ) ( 2 s + 200 ) W ( s ) + sW ( s )
Set K P = 2 and K I = 200 . X (s) = P
= 1 2

s + 20 s + 200 s + 20 s + 200
2 2
1094
TimeResponses:
1095
10-63)
10 10
1 2 3 6 11 6
Consider:
10 6 11 6
Therefore:
0 1 0 0
0 0 1 0
6 11 6 10
1 0 0
As a result:
0 1 0 0
0 0 1 0 1 0 0 0
6 11 6 10
Using MATLAB, well find:

15.4 4.5 0.8
10-64)
Inverted Pendulum on a cart
The equations of motion from Problem 4-21 are obtained (by ignoring all the pendulum inertia term):
x ml&& cos + ml& 2 sin = f

( M + m) &&

x cos + l&&) = 0
ml ( g sin &&
These equations are nonlinear, but they can be linearized. Hence
0
cos 1
sin
x + ml&& = f
( M + m) &&

x + l&&) = 0
ml ( g &&
1096
Or
( M + m) ml &&
x f
ml =
ml mlg
2 &&

Premultiplybyinverseofthecoefficientmatrix
inv([(M+m),m*l;m*l,m*l^2])
ans=
[1/(M+2*m),1/l/(M+2*m)]
[1/l/(M+2*m),(M+m)/m/l^2/(M+2*m)]
ForvaluesofM=2,m=0.5,l=1,g=9.8
ans=
0.33330.3333
0.33331.6667
Hence
x 1/ 3 1/ 3 f
&&
&& = 1/ 3 5 / 3 49 /10

x 1/3*f-49/30
&&
&& = 1/3*f+49/6

Thestatespacemodelis:
x&4 1/3*f-49/30x1
x& = 1/3*f+49/6x
2 1
Or:
1097
x&2 1/3*f+49/6x1
x& = 1/3*f-49/30x
4 1
x&1 0 1 0 0 x1 0
x& 49/6 0 0 0 x2 1/ 3
2 = + f
x&3 0 0 0 1 x3 0

x&4 -49/30 0 0 0 x4 1/ 3
0 1 0 0
49/6 0 0 0
A=
0 0 0 1

-49/30 0 0 0
0
1/ 3
B=
0

1/ 3
C = [1 0 1 0]
D=0
1098
Use ACSYS State tool and follow the design process stated in Example 10-17-1:
1099
TheAmatrixis:
Amat=
01.000000
8.1667000
0001.0000
1.6333000
ans=
s^449/6*s^2
Abar=
0000
0000
002.85770
0002.8577
EigenVectorsare
T=
000.32390.3239
000.92560.9256
1.00001.00000.06480.0648
00.00000.18510.1851
StateSpaceModelis:
a=
x1x2x3x4
x10100
x28.167000
x30001
x41.633000
b=
u1
x10
x20.3333
10100
x30
x40.3333
c=
x1x2x3x4
y11010
d=
u1
y10
ans=
s^449/6*s^2
Transferfunction:
4.441e016s^3+0.6667s^22.22e016s3.267
s^48.167s^2
Pole,ZeroForm:
Zero/pole/gain:
4.4409e016(s+1.501e015)(s+2.214)(s2.214)
s^2(s2.858)(s+2.858)
TheControllabilityMatrix[BABA^2B...]is=
Smat=
00.333302.7222
0.333302.72220
00.333300.5444
0.333300.54440
ThesystemisthereforeNotControllable,rankofSMatrixis=
rankS=
10101
Mmat=
08.166701.0000
8.166701.00000
01.000000
1.0000000
TheControllabilityCanonicalForm(CCF)Transformationmatrixis:
Ptran=
000.33330
0000.3333
3.266700.33330
03.266700.3333
ThetransformedmatricesusingCCFare:
Abar=
01.000000
001.00000
0001.0000
008.16670
Bbar=
0
0
0
1
Cbar=
3.266700.66670
Dbar=
10102
Note incorporating K in Abar:
AbarK=
01.000000
001.00000
0001.0000
k1k28.1667k3k4
SystemCharacteristicequationis:
k4*s^4+(8.1667k3)*s^3k2*sk1=0
From desired poles we have:
>> collect((s-210)*(s-210)*(s+20)*(s-12))
ans =
-10584000+s^4-412*s^3+40500*s^2+453600*s
Hence: k1=10584000, k2=40500, k3=412+8.1667and k4=1
10-65) If 3 and 0.707, then 1.414. The 2nd order desired characteristic equation of the
system is:
2 2 0 1
On the other hand:
0 1
6 5
where the characteristic equation would be:
5 6 0 (2)
Comparing equation (1) and (2) gives:
5 2
6 2
10103
which means 4 and 3
10-66) Using ACSYS we can convert the system into transfer function form.

1) Activate MATLAB
3) Type in Acsys
8) Next verify Controlability and find the A matrix
9) Follow the design procedures in section 10-17 (pole placement)
10104
10105
The A matrix is:
Amat =
-1 -2 -2
0 -1 1
1 0 -1
ans =
s^3+3*s^2+5*s+5
Abar =
-0.6145 + 1.5639i 0 0
0 -0.6145 - 1.5639i 0
0 0 -1.7709
Eigen Vectors are
T=
-0.8074 -0.8074 -0.4259
0.2756 + 0.1446i 0.2756 - 0.1446i -0.7166
-0.1200 + 0.4867i -0.1200 - 0.4867i 0.5524
State-Space Model is:
a=
x1 x2 x3
x1 -1 -2 -2
x2 0 -1 1
x3 1 0 -1
10106
b=
u1
x1 2
x2 0
x3 1
c=
x1 x2 x3
y1 1 1 1
d=
u1
y1 0
Continuous-time model.
ans =
s^3+3*s^2+5*s+5
Transfer function:
3 s^2 + 7 s + 4
---------------------
s^3 + 3 s^2 + 5 s + 5
Pole, Zero Form:
Zero/pole/gain:
3 (s+1.333) (s+1)
---------------------------------
(s+1.771) (s^2 + 1.229s + 2.823)

10107
Smat =
2 -4 0
0 1 0
1 1 -5
The system is therefore Controllable, rank of S Matrix is =
rankS =
Mmat =
5 3 1
3 1 0
1 0 0
Ptran =
-2 2 2
3 1 0
3 4 1
The transformed matrices using CCF are:
Abar =
0 1.0000 0
0 0 1.0000
-5.0000 -5.0000 -3.0000
10108
Bbar =
Cbar =
4 7 3
Dbar =
Using Equation (10-324) we get:
sI ( A BK ) = s 3 + (3 + k3 ) s 2 + (5 + k2 ) s + (5 + k1 ) = 0
Using a 2nd order prototype system, for 5, then 1. For overshoot of 4.33%, 0.707.
nd
Then the desired 2 order system will have a characteristic equation:
s 2 + 2n s + n 2 = s 2 + 2 s + 2 = 0
The above system poles are: s1,2 = 1 j
One approach is to pick K=[k1 k2 k3] values so that two poles of the system are close to the desired
second order poles and the third pole reduces the effect of the two system zeros that are at z=-1.333 and
z=-1. Lets set the third pole at s=-1.333. Hence
(s+1.333)*(s^2+2*s+2)=s^3+3.33*s^2+4.67*s+2.67
and K=[-2.37 -0.37 0.33].
Y 3( s + 1)
= 2
R s + 2s + 2
Use ACSYS control tool to find the time response. First convert the transfer function to a unity feedback
system to make compatible to the format used in the Control toolbox.
3( s + 1)
G=
s2 s 1
10109
Overshoot is about 2%. You can adjust K values to obtain alternative results by repeating this process.
10-67) a) According to the circuit:
If , , then
10110
or
2 1
Therefore:
2 1
1 1
0
0
b) Uncontrollability condition is:
According to the state-space of the system, C is calculated as:
2 1
As 0, because , then the system is controllable
c) Unobservability condition is:
According to the state-space of the system, C is calculated as:
Since det H 0, because R 0 or L , then the system is observable.
10111
d) The same as part (a)
1 1
0
1
0
1 1
1
For controllability, we define G as:
1 1
If ,and then 0, which means the system is not controllable.
For observability, we define H as:
1
1
1
If , then 0, which means the system is not observable.
10112

Farid Golnaraghi, Benjamin C. Kuo-Automatic Control Systems, 9th Edition - Solutions Manual (2009)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Farid Golnaraghi, Benjamin C. Kuo-Automatic Control Systems, 9th Edition - Solutions Manual (2009)

Uploaded by

Copyright:

Available Formats

Ninth Edition

Farid Golnaraghi Benjamin C. Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

'Generated transfer function:'

'Generated transfer function:'

'Generated transfer function:'

'Generated transfer function:'

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

2-4) Mathematical representation:

In all cases substitute and simplify. The use MATLAB to verify.

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

'Generated transfer function:'

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

'Generated transfer function:'

'Generated transfer function:'

'Generated transfer function:'

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

lim 0 G ( j ) = 5; = tan 1 1 = 90o

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

zeta=0.5; %asuuming a value for zeta <1

wn=2*pi*10 %asuuming a value for wn

zeta=1.5; %asuuming a value for zeta >1

T=3.5 %assuming value for parameter T

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

Steps for plotting |G|:

(1) For < 0.1, asymptote is

Break point: = 0.5

Steps for plotting G

b) Lets convert the transfer function to the following form:

Steps for plotting |G|:

(1) Asymptote: < 1 |G(j)| 2.5 /

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

(2) n =2 and = 0.1 for second-order pole

Steps for plotting G(j):

c) Convert the transfer function to the following form:

(1) the breakpoint: = 1 and slope is zero

Steps for plotting G:

(1) phase starts from -180o due to

Steps for plotting the |G|:

(1) Asymptote for <1 is zero

(2) Breakpoint: = 1, slope = -1 or -10 dB/decade

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

Steps for plotting G:

at = 5. For > 5, the phase remains at -180o.

'Generated transfer function:'

'Generated transfer function:'

'Generated transfer function:'

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

'Generated transfer function:'

'Generated transfer function:'

'from pade approzimation:'

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

AutomaticControlSystems,9thEdition Chapter2Solutions Golnaraghi,Kuo

2-10) We know that:

wn=2pi10 %asuuming a value for wn

% f2 = (cos(2t))^2 = 1-(sin(2t))^2 ===> L(f2)=1/s-L(f1) ===>

f7=(2+2spade(exp(-1s),1)+4pade(exp(-2s),1))/(s^2+3s+2) %using Pade command

f7n=(-2s^3+6s+12)/(s^4+6s^3+13s^2+12*s+4) %generating sumbolic function for

Eq=5(pade(exp(-2s),1))/(s^2+s+1); %Pade approximation oreder 1 used