Professional Documents
Culture Documents
Farid Golnaraghi, Benjamin C. Kuo-Automatic Control Systems, 9th Edition - Solutions Manual (2009)
Farid Golnaraghi, Benjamin C. Kuo-Automatic Control Systems, 9th Edition - Solutions Manual (2009)
l
ua
an
M
ns
tio
lu
So
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
clear all;
s = tf('s')
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
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)
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)
Gb=1/(1+2*zeta*s/wn+s^2/wn^2)
figure(2)
nyquist(Gb)
%Part(c)
wn=2*pi*10
Gc=1/(1+2*zeta*s/wn+s^2/wn^2)
figure(3)
nyquist(Gc)
%Part(d)
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) =
. .
(1) = -90o
0: 90
(2) =
0
0: 0
.
(3) =
.
: 90
.
0: 90
.
(4) =
.
0
.
G(j) = . G(s) =
.
216
(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
0.01 1
0.01 1
9
for term , slope is -2 (-40 dB/decade) and passes through | | 1
(2) G(j)| =1 = 0
(3) G(j)| = 2 = -180o
d) G(j) =
(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
(1) The phase starts at 0o and falls -1 or -20 dB/decade at = 0.2 and approaches -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
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
1
2
2
sG s g 0 g t
sG s g 0
223
Which means:
2-13) Consider:
224
1
2
Then
Where
therefore:
1
G s G s
2
= = sG(s) + g(0)
225
As
Therefore: 0
As
Therefore
0 0
which means:
215)
MATLABcode:
clear all;
syms t
s=tf('s')
f1 = (sin(2*t))^2
L1=laplace(f1)
226
f3 = (cos(2*t))^2
L3=laplace(f3)
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)
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
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)
(a) g (t ) = 5te 5t u s (t )
Answer:5/(s+5)^2
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
Answer:2/(s^2+16)
(e) g ( t ) = e 5 kT
( t kT ) where(t)=unitimpulsefunction
k =0
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)
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)
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')
225)
MATLABcode:
clear all;
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
>> 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 =
1 6 10 5 0
>> b=num*2
b=
2 2 2
>> a=denom
a=
1 6 10 5 0
r=
-0.9889
234
2.5889
-2.0000
0.4000
p=
-3.6180
-1.3820
-1.0000
k=[]
b r r r
= 1 + 2 + ... + n + k
a s + p1 s + p2 s + pn
(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')
236
syms s
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
(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
4 5 2
As X s e Y s , then
240
2 1
4
Then:
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));
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part c
Eq=5*(s+2)/(s^2*(s+1)*(s+5));
[num,den]=tfdata(Eq,'v');
242
[r,p] = residue(num,den)
%Part d
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part e
Eq=100*(s^2+s+3)/(s*(s^2+5*s+3));
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part f
Eq=1/(s*(s^2+1)*(s+0.5)^2);
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part g
Eq=(2*s^3+s^2+8*s+6)/((s^2+4)*(s^2+2*s+2));
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
%Part h
Eq=(2*s^4+9*s^3+15*s^2+s+2)/(s^2*(s+2)*(s+1)^2);
[num,den]=tfdata(Eq,'v');
[r,p] = residue(num,den)
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
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.
(d)Polesareat s = 5, 1 + j , 1 j AllpolesintheLHP.Stable.
(e)Polesareat s = 13387
. , 16634
. + j 2.164, 16634
. j 2.164 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
sr = roots(p)
p=
1 5 2 10
251
sr =
-5.0000
0.0000 + 1.4142i
0.0000 - 1.4142i
(c)
ans =
-3.6180
-1.3820
ans =
-5.0000
-1.0000 + 1.0000i
-1.0000 - 1.0000i
ans =
1.6694 + 2.1640i
1.6694 - 2.1640i
-1.3387
252
-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);
[num,den]=tfdata(Eq,'v');
roots(den)
%Part b
Eq=(s-1)/((s+5)*(s^2+2));
[num,den]=tfdata(Eq,'v');
roots(den)
%Part c
Eq=1/(s^3+5*s+5);
[num,den]=tfdata(Eq,'v');
roots(den)
253
%Part d
Eq=100*(s-1)/((s+5)*(s^2+2*s+2));
[num,den]=tfdata(Eq,'v');
roots(den)
%Part e
Eq=100/(s^3-2*s^2+3*s+10);
[num,den]=tfdata(Eq,'v');
roots(den)
%Part f
Eq=10*(s+12.5)/(s^4+3*s^3+50*s^2+s+10^6);
[num,den]=tfdata(Eq,'v');
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)
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
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.
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
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
ans =
-12.4799 + 9.6566i
-12.4799 - 9.6566i
259
-0.0402
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
2. GotothedirectorycontainingtheACSYSsoftware.
3. Typein
Acsys
4. ThenpressthetransferfunctionSymbolicbutton
270
5. EnterthecharacteristicequationinthedenominatorandpresstheRouthHurwitzpush
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
roots(den_c)
%Part d
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]
Stablesystem>>Norighthandsidepole
Part(d):
RHchart:
[2,11/2,10]
[10,11/2,0]
[22/5,10,0]
[379/22,0,0]
[10,0,0]
Unstablesystemdueto379/22onthe4throw.
2complexconjugatepolesonrighthandside.Allthepolesare:
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.
4complexconjugatepolesonrighthandside.Allthepolesare:
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]
Stablesystem>>Norighthandsidepole
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
2. GotothedirectorycontainingtheACSYSsoftware.
3. Typein
Acsys
283
4. ThenpressthetransferfunctionSymbolicbutton.
284
5. EnterthecharacteristicequationinthedenominatorandpresstheRouthHurwitzpush
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.
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.
Thefrequencyofoscillationis11.832rad/sec.
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
2 2
WhenK=6000,theauxiliaryequationis A( s ) = 30 s + 6000 = 0. Thesolutionis s = 200.
Thefrequencyofoscillationis14.142rad/sec.
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
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
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
UsetheapproachinthisChaptersSection214:
1. ActivateMATLAB
2. GotothedirectorycontainingtheACSYSsoftware.
3. Typein
Acsys
290
4. ThenpressthetransferfunctionSymbolicbutton.
291
5. EnterthecharacteristicequationinthedenominatorandpresstheRouthHurwitzpush
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.
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
2. GotothedirectorycontainingtheACSYSsoftware.
3. Typein
Acsys
4. ThenpressthetransferfunctionSymbolicbutton.
293
5. EnterthecharacteristicequationinthedenominatorandpresstheInverseLaplaceTransform
pushbutton.
----------------------------------------------------------------
----------------------------------------------------------------
294
G(s) =
[ kd kp ]
[------------------------ ------------------------------]
[ 3 3 ]
G(s) factored:
[ kd kp ]
[ 3 3 ]
[ s + 10 s kd + 10 kp - 100 s + 10 s kd + 10 kp - 100]
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
Using ACSYS:
RH =
[ 1/10, 1]
[ 9/10, 9]
[ 9/5, 0]
[ 9, 0]
----------------------------------------------------------------
----------------------------------------------------------------
G(s) =
s + 10
-------------------------
3 2
1/10 s + 9/10 s + s + 9
296
G factored:
Zero/pole/gain:
10 (s+10)
-----------------
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)
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:
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
3-7)
33-8)
36
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
3-9)
3
3-10)
3
3-11)
37
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
3
3-21)
3
3-22)
314
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
33-23)
315
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
3-24)
3
3-25)
3
3-26)
3
3-27)
316
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
33-28)
317
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
33-29)
318
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
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
AutomaticControlSystems,9thEdition
A Chapter3Solution
ns Golnarraghi,Kuo
332)
321
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
[r,p,k]=residue(B,A)
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
AutomaticControlSystems,9thEdition
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
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])
[r,p,k]=residue(B,A)
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:
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
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
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
-------------------------------------------------------------------------------
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
+
1.152e007 s
-------------------------------------------------------------------------------
16 s^21 + 448 s^20 + 5904 s^19 + 49252 s^18 + 294265 s^17 + 1.346e006 s^16
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
+
1.152e007 s
-------------------------------------------------------------------------------
8 s^20 + 308 s^19 + 5270 s^18 + 54111 s^17 + 379254 s^16 + 1.955e006 s^15
337
Chapter 4__________________________________________________________________________
4-1) When the mass is added to spring, then the spring will stretch from position O to position L.
1
2
where y is a displacement from equilibrium position L.
1
2
0
since KL = mg, therefore:
0
41
4-2)
KU
KL
1
2
Potential energy:
As we know , then:
42
1 1 1
2 2 2
By differentiating, we have:
0
b)
c)
1 1 1
2 2 2
1 1
2 2
Then:
1 1
2 2
Or:
43
J
d) G (s) =
(ms + K )
2
Transferfunction:
10
s^2+10
44
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
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:
Stattediagram:(W
Withminimumnumberofinttegrators)
49
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
Then:
410
c)
1
2
d)
0 0 1 0
0 0 0 1 0
0
0 1
0 0
0001 0
4-7)(a)Forceequations:
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
sin
0
sin sin sin cos
sin
cos 1 0
sin 0
cos 1
4-9) a)
412
Fy
m
x(t) Fx
mg
M
f
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
1 1
sin cos
2 2
1 1
sin cos
2 2
414
sin 0
b) As:
Then
sin 0
If we linearize the equation about beam angle = 0, then sin and sin
Then:
415
c) Considering
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)
xlabel( 'Time(sec)');
ylabel('Amplitude');
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
417
Refer to the aircraft dynamics textbooks, the state equations can be written as:
1
2
3
Which gives:
418
For plotting
g state flow diagram, equation
u (4) muust be rewritten as:
or
ml 1/Ms 1 1/s
u y
x
-B/M
-K
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
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:
(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
y + &&
m( && ym ) = Fm
( M m) &&
y = Fm By& Ky
ym = e 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)
xlabel( 'Time(sec)');
ylabel('Amplitude');
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
2 1
0 2
2
0
0 0 1 0
0 0 0 1 0
2 0
0 0 1
0 0 0
b)
Which gives:
and
Therefore:
2 2
431
where
as a result:
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
1
4 1
i1 i2 RL
Vin VC1 VC 2
Then we have:
When
If RL >>R, then
433
Therefore:
can be obtained by substituting above expressions into the first equation of the state variables
of the system.
2
Since the electric force is opposes the motion of the plates, then the equation of the motion is written
as:
434
Since , then :
b) As then
If 1
435
If 1
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.
436
,
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:
Also, the current is changing with the rate of changes in displacement. It means:
i Bx
So:
1
2
Therefore:
V s 2
tan 1
F s B
437
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
,
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:
438
or
1
1 1
1
Therefore:
439
As
10
1
then:
10
1
1
which gives:
1
10 1 10
Then:
Also:
0
Then:
1
1 0
As a result:
440
1 1
1
Or
1
0
As a result:
or
As a result:
1
1
4-32) The heat flow-in changes with respect to the electric power as:
where Kf is the heat flow coefficient between actuator and air, T1 and T2 are temperature of
actuator and ambient.
441
1 1 1
Therefore:
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
ln
, 3
2
: 4
: 5
: 6
, 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.
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
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
1 1
2 2
The tottal kinetic en
nergy is:
445
Therefore:
2
As a result:
2
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:
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
| |
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)
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.
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
As a result
0 0
sin
cos sin
cos
0
cos sin
where L is the length of lever and r is the offset from the center of rotation.
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:
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
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 )
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
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
Thesam
meresultsasab
boveareobtained.
4-550)(a)Systemequations:
(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
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
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 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
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
466
where
2
The resulted state space equations are:
2
2
467
4-60) As the base is not moving then the model can be reduced to:
Therefore:
4-61)
V2
| |
A
C
g
468
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)
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
(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
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
________________________________________________________________________________
u s ( t ) K p = 1000 1 1001
tu s ( t ) K v = 0
2
t us (t ) / 2 K a = 0
________________________________________________________________________________
u s ( t ) K p = 0
tu s ( t ) K v = 1 1
2
t us (t ) / 2 K a = 0
________________________________________________________________________________
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.
________________________________________________________________________________
u s ( t ) K p = 0
tu s ( t ) K v = 1 1
2
t us (t ) / 2 Ka = 0
________________________________________________________________________________
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
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
Zero: s = -1
5-8)
5( s + 1)
s ( s + 2)( s + 3)
5 1
2 3
58
5-10)
4( s + 1)
s 2 ( s + 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
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
. . . .
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
(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):
513
517YoumayuseMATLABinallRouthHurwitzcalculations.
1. ActivateMATLAB
2. GotothedirectorycontainingtheACSYSsoftware.
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):
515
518
2
1
Percent maximum overshoot = 0.25 = e
Thus
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
Y ( s) 117
= 2
R( s ) s + 15.3s + 117
UnitstepResponse:
y=0.1att=0.047sec.
y=0.9att=0.244sec.
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
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
Y ( s) 313
= 2
R( s ) s + 20.88 s + 314.5
UnitstepResponse:
y=0.1whent=0.028sec.
y=0.9whent=0.131sec.
y max = 11
. (10% max. overshoot )
518
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
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
Y ( s) 802.59
= 2
R( s ) s + 25.84 s + 802.59
UnitstepResponse:
y=0.1whent=0.0178sec.
y=0.9whent=0.072sec.
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
Y ( s) 202.5
= 2
R( s ) s + 20.1s + 202.5
UnitStepResponse:
Wheny=0.5,t=0.1005sec.
y max = 1043
. ( 4.3% max. overshoot )
520
523ClosedLoopTransferFunction: 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 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
Y ( s) 715
= 2
R( s ) s + 3155
. s + 715
UnitStepResponse:
y=0.5whent=0.0505sec.
Thus, t d = 0.0505sec.
y max = 11007
. (10.07% max. overshoot )
521
524ClosedLoopTransferFunction: 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=0.2, = 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
Y ( s) 15737.7
= 2
R( s ) s + 114.41s + 15737.7
UnitstepResponse:
y=0.5whent=0.0101sec.
Thus, t d = 0.0101sec.
522
525ClosedLoopTransferFunction: CharacteristicEquation:
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)
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.
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)
2
. n 5
118 n
Kt = = 0.246 K= = 470.63
500 25
SystemTransferFunction:
Y ( s) 11765.74
= 2
R( s ) s + 128 s + 11765.74
UnitStepResponse:
y(t)reaches1.05andneverexceeds
thisvalueatt=0.048sec.
0.912 n 5
Kt = = 1.27
500
SystemTransferFunction:
Y ( s) 492453
= 2
R( s ) s + 640 s + 492453
UnitStepResponse:
524
y(t)reaches1.05andnever
exceedsthisvalueatt=0.0074sec.
thanthecalculatedvalueof0.01sec.
525
527ClosedLoopTransferFunction: 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)
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
SystemTransferFunction:
Y ( s) 1012.2
= 2
R( s ) s + 44.986 s + 1012.2
UnitStepResponse:Theunitstepresponsereaches0.95att=0.092sec.whichisthemeasured t s .
526
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
ForwardPathTransferFunction:
69.58
G( s) =
s( s + 16
. )( s + 30)
UnitStepResponse:
y=0.1whent=0.355sec.
y=0.9whent=1.43sec.
RiseTime:
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
(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,
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)
(b)For = 1, 2 n = 100 K D .
(c)Seeparameterplaneinpart(g).
531
(d)Seeparameterplaneinpart(g).
2
(e)Parabolicerrorconstant K a = 1000 sec
(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
. . . . .
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 )
0.125
2 5.42
. .
5.49
.
19.88
535
5-35) a)
6.5
1
6.5 2
.
1
6.5 1
. .
, , ,
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
1
1
537
where n = 1, and 2 n = 1 = 1 / 2
sin 1
2 3
1 sin
3 2
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);
[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:
, , ,
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
sin 1
1
b)
sin 1
1
c)
1
1 sin 1
1
where = cos-1
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
sin
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.
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
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
UnitStepResponse:N=10
SecondorderSystem ThirdorderSystem
549
543UnitStepResponses:
derivativecontrolorahighpassfilter.
550
544UnitStepResponses
1
Theeffectofaddingthepoleat s = toG(s)istoincreasetherisetimeandtheovershoot.Thesystemis
Tp
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.
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
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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
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
k
K Kf
+1
Js + a s
Accordingly,
and
555
1 1
: lim
1
1
lim
1
lim
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)
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 )
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
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);
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
Transfer function:
10 s + 562.5
-----------------
s^2 + 64 s + 1288
559
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
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
2 1 6 2
2 13 3.25
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)
0.2
0.109
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
. . .
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
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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
564
. .
As , then, 3.62
. .
1 3
3 3
Therefore:
3 3 2
2 3
2 3
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);
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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
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
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
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);
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
Zero/pole/gain:
2000
-------------------------
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
Theforwardpathtransferfunctionis Thecontrollertransferfunctionis
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);
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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)
Matchwitha2ndorderprototypesystem
solve(10+1000*Kd100)
ans=
9/100
90
Thus KD = = 0.09
1000
Usethesameprocedureforotherparts.
131.4
2 n = 10 + 1000 K D = 2 0.707 100 = 141.4 Thus KD = = 0.1314
1000
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
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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
Zero/pole/gain:
360.9 (s+1000)
s (s+361.2)
Zero/pole/gain:
360.9 (s+1000)
Zero/pole/gain:
541.35 (s+666.7)
s (s+361.2)
Zero/pole/gain:
541.35 (s+666.7)
Zero/pole/gain:
721.8 (s+500)
s (s+361.2)
Zero/pole/gain:
721.8 (s+500)
581
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)
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
xlabel('Time(secs)')
ylabel('y(t)')
title('Unit-step responses of the system')
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(%)
0.05 2.721 0
0.06 3.039 0
0.10 4.317 0
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)
ForwardpathTransferFunction:
200 ( 0.05 + K D s )
G ( s ) =
s ( s + 1)( s + 10)
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
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
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)
2
125 + 20 3
KP = = 15765
.
100
(c)RootContours:
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)
(a)ForwardpathTransferFunction:
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
8. GotothedirectorycontainingtheACSYSsoftware.
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];
[numCL,denCL]=cloop(num,den);
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
597
(a)ForwardpathTransferFunction:
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.
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
ForwardpathTransferFunction:
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];
[numCL,denCL]=cloop(num,den);
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];
[numCL,denCL]=cloop(num,den);
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];
[numCL,denCL]=cloop(num,den);
GCL=tf(numCL,denCL)
step(GCL)
Transfer function:
100 s^2 + 3000 s + 10000
------------------------------
s^3 + 110 s^2 + 3100 s + 10000
5101
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.
xc
mc
mc
ks cs
ks cs x
xw m
mw
mw k c
kw kw cw
y y
cw
Figure 4-84 Quarter car model realization: (a) quarter car, (b) 2 degree of
Theequationofmotionofthesystemisdefinedasfollows:
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:
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
5104
Transfer function:
-s
---------------------
s^3 + 1.414 s^2 + 2 s
5105
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
5106
1
2
Potential energy:
As we know , then:
1 1 1
2 2 2
By differentiating, we have:
0
b)
c)
5107
1 1 1
2 2 2
1 1
2 2
Then:
1 1
2 2
Or:
J
d) G (s) =
(ms + K )
2
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
5109
Kp =
0.1000
Transfer function:
0.5
--------------
25 s^2 + 100.5
ans =
0 + 2.0050i
0 - 2.0050i
5110
a)
y2 y1
K ( y1 y 2 )
K ( y1 y 2 )
mgy& 2 Mgy&1
Then:
The output equation can be the velocity of the engine, which means
c)
5111
1
2
KP
Zref_FZ
1
2
1
2 1
x(t)
F
L
M
Here is an alternative representation including friction (damping) . In this case the angle is
measured differently.
5112
1)
2)
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 ) )
A( K p + KDs)
= 2
R (s + K D s + A ( K p B )
5113
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];
den =[1 Kd A*(Kp-B)];
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
For a better design, and to meet rise time criterion, use Example 5-
11-1.
5116
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. +15 V input:
c. 10 V input:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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:
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. +15 V:
c. 10 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
9. Sine input to SIMLab and Virtual Lab (1 V. amplitude, and 0.5, 5, and 50 rad/sec frequencies)
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. 5 rad/sec (SIMLab):
c. 50 rad/sec (SIMLab):
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. 5 rad/sec (SIMLab):
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
c. 50 rad/sec (SIMLab):
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
a. +5 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. +15 V:
c. -10 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. +15 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
c. -10 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
By decreasing the C.F. from 1 to 0.2, the final value of the speed increases by a factor of 5.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
b. +15 V:
c. 10 V:
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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.
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.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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.
11.2
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-4
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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).
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-10
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
a) 1 radian.
b) 1.203 radians.
c) 0.2215 seconds.
6-11
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
As the derivative gain increases, overshoot decreases, but rise time increases.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-14
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-15
There could be many possible answers for this problem. One possible answer would be
Kp= 100
Ki= 10
Kd= 1.4
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-16
0.1 Hz
0.2 Hz
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
0.5 Hz
1 Hz
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
2Hz
5Hz
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
10Hz
50Hz
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-18
Considering fast response time and low overshoot, Kp=1 is considered to be the best value.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
As Kd value increases, the overshoot decreases and the rise time increases.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-20 From the Experiment Menu select the Open Loop Sine Input option.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
Double click on the Sine wave block and choose the input values.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
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
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-21)
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
Run simulation, and repeat the process for different gain values. Observe the steady state value change
with K.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-22)
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
Run simulation, and repeat the process for different gain values.
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-25) a)
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
c) A sample response
d)
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
Automatic Control Systems, 9th Edition Chapter 6 Solutions Golnaraghi, Kuo
6-1
AutomaticControlSystems,9thEdition Chapter7Solutions Golnaraghi,Kuo
Chapter 7
7-1 (a) P ( s ) = s 4 + 4 s 3 + 4 s 2 + 8 s Q( s ) = s + 1
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
Intersect of Asymptotes:
4.7913 0.2087 ( 1)
1 = = 2
31
(c) P ( s ) = s 2 3 2
Q( s ) = s + 3s + 2 s + 8
71
(d) P ( s ) = s 3 + 2 s 2 + 3 s ( )
Q( s) = s 1 ( s + 3)
2
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
Intersect of Asymptotes:
1 1 ( 15)
. ( 15)
. 1
1 = =
5 2 3
(f) P ( s ) = s 4 + 2 s 2 + 10 Q( s ) = s + 5
Intersect of Asymptotes:
10398
. 10398
. + 10398
. + 10398
. ( 5) 5
1 = =
4 1 3
72
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
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
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)
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
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)
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
|s| > |zi| for i = 1,2,,m and |s| > |pi| for i = 1,2,,n
therefore G(s)H(s)
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
since the intersect of (n-m) asymptotes lies on the real axis of the s-plane and is
real, therefore
710
j
j
711
K > 0: 1 2 3 + 4 = 180o
1 90 45 + 90 = 180
o o o o
1 = 135
o
1 = 45
o
1 2 3 + 4 = 180
o
K > 0:
1 135 90 + 90 = 0
o o o o
K < 0:
1 = 135
o
712
1 2 3 + 4 = 180
o
K > 0:
1 135 90 45 = 180
o o o o
1 = 90
o
1 2 3 4 = 180
o
K > 0:
1 = 180
o
713
1 + 6 2 3 4 5 = 360
o
K < 0:
1 = 108.435
o
714
7-7) a) G(s)H(s) =
we know that
therefore
b) Similarly:
G(s)H(s) =
Therefore:
As a result, =180
715
Arrival angles at
7-9)(a) (b)
(c) (d)
716
AutomaticControlSystems,9thEdition Chapter7Solutions Golnaraghi,Kuo
717
718
7-10) The breaking points are on the real axis of and must satisfy
or
5 4 3 2
7-11) (a) Breakaway-point Equation: 2 s + 20 s + 74 s + 110 s + 48 s = 0
6 5 4 3 2
(b) Breakaway-point Equation: 3 s + 22 s + 65 s + 100 s + 86 s + 44 s + 12 = 0
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
6 5 4 3 2
(d) Breakaway-point Equation: s 8 s 19 s + 8 s + 94 s + 120 s + 48 = 0
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
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
IntersectofAsymptotes:
0 1 3 4
1 = = 2
4
3 2
BreakawaypointEquation: 4 s + 24 s + 38 s + 12 = 0
RootLocusDiagram:
721
712(c)
K( s + 4)
G( s) H ( s) = 2 2
s ( s + 2)
IntersectofAsymptotes:
0 + 0 2 2 ( 4 )
1 = = 0
4 1
BreakawaypointEquation:
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 )
IntersectofAsymptotes:
0 1 j 1 j ( 2 )
1 = = 0
31
3 2
BreakawaypointEquation: 2s + 8s + 8s + 4 = 0
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
IntersectofAsymptotes:
0 1 j 1 j ( 5)
1 = = 15
.
31
BreakawaypointEquation:
3 2
2 s + 17 s + 20 s + 10 = 0
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
IntersectofAsymptotes:
0 1 j 1+ j 4
1 = = 15
.
4
3 2
BreakawaypointEquation: 4 s + 18 s + 20 s + 8 = 0
725
712(g)
2
K( s + 4)
G( s) H ( s) = 2 2
s ( s + 8)
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)
IntersectofAsymptotes:
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
IntersectofAsymptotes:
8 8 ( 4 ) ( 4 )
1 = = 4
42
5 4 3 2
BreakawaypointEquation: s + 20 s + 128 s + 736 s + 1280 s = 0
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 )
Asymptotes: K>0: 90 o , 270 o K<0:
o o
0 , 180
2 + 2
IntersectofAsymptotes: 1 = = 0
42
6 4 2
BreakawaypointEquation: s 8 s 24 s = 0
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
IntersectofAsymptotes:
1 + 1
1 = = 0
42
5 3
BreakawaypointEquation: s 2s 9s = 0
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)
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 )
735
IntersectofAsymptotes:
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
num=[0 1]; 2
Imaginary Axis
-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]);
den=conv(den,[1 2]);
den=conv(den,[1 2]); -10
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
739
Root Locus
5
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
den=conv(den,[1 8]); 2
Imaginary Axis
den=conv(den,[1 8]);
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
num=[0 1]; 5
Imaginary Axis
den=conv([1 0],[1 0]);
den=conv(den,[1 8]); 0
den=conv(den,[1 8]); -5
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
741
Root Locus
0.25
0.2
0.15
0.1
0.05
Imaginary Axis 0
-0.25
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
Real Axis
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
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
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
den=conv(den,[1 0]);
0
den=conv(den,[1 100]);
den=conv(den,[1 200]);
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
7-14)(a) Q( s ) = s + 5 ( )
P ( s ) = s s + 3s + 2 = s ( s + 1)( s + 2)
2
IntersectofAsymptotes:
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
IntersectofAsymptotes:
1 ( 3)
1 = = 1
31
3 2
BreakawaypointEquation: s + 5s + 3s + 3 = 0
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
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
750
714(e) ( )
Q( s ) = s 1 ( s + 2 )
2
(
P( s) = s s + 2s + 2
2
)
751
3 2
BreakawaypointEquation: 6 s + 12 s + 8 s + 4 = 0
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
754
714(g) Q( s ) = s + 4 s + 5
2
P( s ) = s
2
(s 2
)
+ 8s + 16
IntersectofAsymptotes:
8 ( 4 )
1 = = 2
42
5 4 3 2
BreakawaypointEquation: s + 10 s + 42 s + 92 s + 80 s = 0
755
714(h) (
Q( s ) = s 2
2
) ( s + 4) (
P( s) = s s + 2s + 2
2
)
756
BreakawayPoints: 2,6.95
757
714(i) Q( s ) = ( s + 2 )( s + 0.5)
2
P( s) = s s 1
4 3 2
BreakawaypointEquation: s + 5s + 4 s 1 = 0
758
759
714(j) Q( s) = 2s + 5 P(s) = s
2
(s 2
)
+ 2s + 1 = s
2
( s + 1)
2
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
%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
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-6 -5 -4 -3 -2 -1 0 1
Real Axis
763
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
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
1.5
0.5
Imaginary Axis
-0.5
-1
-1.5
-6 -5 -4 -3 -2 -1 0 1 2
Real Axis
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
Asymptotes angle:
Therefore,
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
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
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
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
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-10 -5 0 5
Real Axis
770
Root Locus
6
2
Imaginary Axis
-2
-4
-6
-6 -5 -4 -3 -2 -1 0 1 2 3 4
Real Axis
IntersectofAsymptotes:
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)
771
IntersectofAsymptotes:
772
0 2 5 10
1 = = 4.25
4
3 2
BreakawaypointEquation: 4 s + 51s + 160 s + 100 = 0
773
4 3 2
BreakawaypointEquation: s + 4 s + 10 s + 300 s + 500 = 0
774
775
o o
716(d)K>0: 90 , 270
IntersectofAsymptotes:
2 2 5 6
1 = = 7.5
42
776
%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)
Root Locus
15
10
System: G_a
Gain: 19
Pole: -0.584 + 0.589i
5 Damping: 0.704
Overshoot (%): 4.43
Imaginary Axis
System: G_a
Gain: 19
-5 Pole: -0.584 - 0.589i
Damping: 0.704
Overshoot (%): 4.43
Frequency (rad/sec): 0.829
-10
-15
-20 -15 -10 -5 0 5 10 15
Real Axis
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
3
1
Imaginary Axis
-1
-2
-3
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2
Real Axis
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
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-15 -10 -5 0 5 10 15 20
Real Axis
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
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
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
IntersectofAsymptotes:
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
IntersectofAsymptotes:
0 1 3 5
1 = = 2.25
4
3 2
BreakawaypointEquation: 4 s + 27 s + 46 s + 15 = 0
784
Angle of asymptotes:
Then and
d) Poles:
Angle of asymptotes:
breakaway points:
Angle of asymptotes:
785
breakaway points:
f) Poles:
Angles of asymptotes:
breakaway point:
%a)
num_G_a=1;
den_G_a=s*(s+10)*(s+20);
G_a=num_G_a/den_G_a;
figure(1);
rlocus(G_a)
%b)
num_G_b= 1;
den_G_b=s*(s+1)*(s+3)*(s+5);
G_b=num_G_b/den_G_b;
figure(2);
rlocus(G_b)
%c)
num_G_c=(s-0.5);
den_G_c=(s-1)^2;
G_c=num_G_c/den_G_c;
figure(3);
rlocus(G_c)
%d)
786
num_G_d=1;
den_G_d=(s+0.5)*(s-1.5);
G_d=num_G_d/den_G_d;
figure(4);
rlocus(G_d)
%e)
num_G_e=(s+1/3)*(s+1);
den_G_e=s*(s+1/2)*(s-1);
G_e=num_G_e/den_G_e;
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
-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
-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
-0.5
-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
-0.5
-1
-1.5
-5 -4 -3 -2 -1 0 1 2
Real Axis
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
%a)
n=1;
num_G_a= 1;
den_G_a=(s+4)^n;
G_a=num_G_a/den_G_a;
figure(n);
rlocus(G_a)
%b)
n=2;
num_G_b= 1;
den_G_b=(s+4)^n;
G_b=num_G_b/den_G_b;
figure(n);
rlocus(G_b)
%c)
n=3;
num_G_c= 1;
den_G_c=(s+4)^n;
G_c=num_G_c/den_G_c;
figure(n);
rlocus(G_c)
%d)
n=4;
num_G_d= 1;
den_G_d=(s+4)^n;
G_d=num_G_d/den_G_d;
figure(n);
rlocus(G_d)
%e)
n=5;
num_G_e= 1;
den_G_e=(s+4)^n;
G_e=num_G_e/den_G_e;
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
o o
Asymptotes: Kt > 0: 90 , 270
IntersectofAsymptotes:
25 0
1 = = 12.5
31
3 2
BreakawaypointEquation: s + 12.5 s 50 = 0
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
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
IntersectofAsymptotes:
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
IntersectofAsymptotes:
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
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
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
o
Asymptotes: J L = 0: 180
4 3 2
BreakawaypointEquation: 2.05 s 479 s 12532 s 37782 s = 0
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
150
100
50
Imaginary Axis
-50
-100
-150
-400 -350 -300 -250 -200 -150 -100 -50 0 50
Real Axis
7102
o
Asymptotes: K>0: 180
4 3 2
BreakawaypointEquation: s + 16 s + 46 s 15 = 0
7103
728(b) P( s ) = s ( s 2 + 10 s + 29 ) Q ( s ) = 10( s + 3)
IntersectofAsymptotes:
0 10 ( 3)
1 = = 3.5
31
3 2
BreakawaypointEquation: 20 s + 190 s + 600 s + 870 = 0
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
20
15
10
5
Imaginary Axis
-5
-10
-15
-20
-6 -5 -4 -3 -2 -1 0 1
Real Axis
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
7107
o o o
Asymptotes: N>0: 60 , 180 , 300
IntersectofAsymptotes:
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
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
s = tf('s')
%a)
A=50;
K0=50;
num_G_a = 250;
den_G_a = 0.06*s*(s + 12.5)*(A*s+K0);
G_a = num_G_a/den_G_a;
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
7112
Root Locus
10
2
Imaginary Axis
-2
-4
-6
-8
-10
-35 -30 -25 -20 -15 -10 -5 0 5
Real Axis
o o o
Asymptotes: N>0: 60 180 300
IntersectofAsymptotes:
0 1 12.5
1 = = 4.5
3
2
BreakawaypointEquation: 3 s + 27 s + 12.5 = 0
BreakawayPoints:(RL) 0.4896
7113
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
Q( s ) = 0.01s( s + 12.5)
4 3 2
BreakawaypointEquation: 0.01s + 0.15 s + 1.5625 s 16.67 s 104.17 = 0
7116
s = tf('s')
%a)
A=100;
K0=100;
num_G_a = 250;
den_G_a = 0.06*s*(s + 12.5)*(A*s+K0);
G_a = num_G_a/den_G_a;
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)
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
8
2
Imaginary Axis
-2
-4
-6
-8
-30 -25 -20 -15 -10 -5 0 5
Real Axis
7119
Angle of asymptotes:
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)
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
IntersectofAsymptotes:
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
IntersectofAsymptotes:
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
;
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
rlocus(GH_b)
Root locus diagram, part (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:
Breakaway points:
Which means:
b)
S2 1 2+2k
7123
S 3-k 0
S0 (3-k)(2+2k)
As a result:
c) In this problem, e Ts term is a time delay. Therefore, MATLB PADE command is used for pade
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)
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
7126
Root Locus
15
10
5
Imaginary Axis
-5
-10
-15
-20 -15 -10 -5 0 5 10
Real Axis
o o
Asymptotes: K L > 0: 90 , 270
IntersectofAsymptotes:
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
IntersectofAsymptotes:
BreakawaypointEquation:
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
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);
G_a=num_G_a/den_G_a;
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);
G_b=num_G_b/den_G_b;
figure(2);
rlocus(G_b)
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 )
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
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);
G_a = num_G_a/den_G_a;
figure(1);
rlocus(G_a)
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
IntersectofAsymptotes:
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
IntersectofAsymptotes:
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
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;
den_GH=s^3+alpha*s^2;
GH=num_GH/den_GH;
figure(2);
rlocus(GH)
%(c)
alpha=9
num_GH= s+1;
den_GH=s^3+alpha*s^2;
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
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
7138
7139
745(d)Twobreakawaypoints: > 3 :
H 1 (s) e st
where
Now designing a controller is similar to the designing a controller for any unity feedback system.
7140
be . Let
Thentheangleconditionsontherootlociare:
o
1 = 2 = 3 = 4 = 90 =0
o
Thus,
Therootlocishownin(b)arethecorrectones.
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)
num_G_a= 5; -30
den_G_a=s*(s+6.54); -40
G_a=num_G_a/den_G_a;
-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);
G_c=num_G_c/den_G_c;
81
-60
-80
-100
0
-45
Phase (deg)
-90
System: CL_b
-135 Frequency (rad/sec): 4.62
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)
-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
where
and
Therefore:
As a result:
(c) M r = 4.17 (12.4 dB) r = 6.25 rad / sec BW = 9.18 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)
2
1 1 0.416 + 2.917
Mr = = 105
. tr = = 0.1 sec
2 1
2 n
((1 2 )+ )
1/ 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
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
8-8) (a)
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);
G_a=num_G_a/den_G_a;
CL_a = G_a/(1+G_a)
BW = bandwidth(CL_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
Frequency (rad/sec): 1.5
Magnitude (dB): -3.01
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:
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:
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
_________________________________________________________________
813)
T BW(rad/sec) Mr
_________________________________________________________________
0 1.14 1.54
_________________________________________________________________
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
Frequency (rad/sec): 1.5
Magnitude (dB): -3.01
Magnitude (dB)
Notes: -40
-60
812
815(a)
20
L( s ) = P = 1, P = 0
s(1 + 0.1s )(1 + 0.5s )
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
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);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
813
(b)
10
L( s ) = Basedontheanalysisconductedinpart(a),theintersectofthenegative
s(1 + 0.1s )(1 + 0.5s )
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);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
814
(c)
100(1 + s )
L( s ) = P = 1, P = 0.
s(1 + 0.1s )(1 + 0.2 s )(1 + 0.5s )
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
815
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);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
816
(d)
10
L( s ) = 2
P = 2 P = 0
s (1 + 0.2 s )(1 + 0.5s )
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
originwhere = .
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);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
817
815(e)
3( s + 2)
L( s ) = P = 1 P = 2
(
s s + 3s + 1
3
)
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
Theclosedloopsystemisunstable.Thecharacteristicequationhastworootsintherighthalfsplane.
MATLAB code:
818
s = tf('s')
%e)
figure(1);
num_G_a= 3*(s+2);
den_G_a=s*(s^3+3*s+1);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
815(f)
0.1
L( s ) = P = 1 P = 0
(
s ( s + 1) s + s + 1
2
)
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
819
MATLAB code:
s = tf('s')
%f)
figure(1);
num_G_a= 0.1;
den_G_a=s*(s+1)*(s^2+s+1);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
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
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
Theclosedloopsystemisunstable.Thecharacteristicequationhastworootsintherighthalfsplane.
820
MATLAB code:
s = tf('s')
%g)
figure(1);
num_G_a= 100;
den_G_a=s*(s+1)*(s^2+2);
G_a=num_G_a/den_G_a;
nyquist(G_a)
NyquistPlotof L( j ):
815(h)
10( s + 10)
L( s ) = P = 1 P = 0
s( s + 1)( s + 100)
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
therealaxis,exceptattheorigin.
Theclosedloopsystemisstable.
MATLAB code:
s = tf('s')
%h)
figure(1);
num_G_a= 10*(s+10);
den_G_a=s*(s+1)*(s+100);
G_a=num_G_a/den_G_a;
nyquist(G_a)
822
NyquistPlotof L( j ):
8-16
MATLAB code:
s = tf('s')
%a)
figure(1);
num_G_a= 1;
den_G_a=s*(s+2)*(s+10);
G_a=num_G_a/den_G_a;
nyquist(G_a)
%b)
figure(2);
num_G_b= 1*(s+1);
den_G_b=s*(s+2)*(s+5)*(s+15);
G_b=num_G_b/den_G_b;
nyquist(G_b)
%c)
figure(3);
num_G_c= 1;
den_G_c=s^2*(s+2)*(s+10);
G_c=num_G_c/den_G_c;
nyquist(G_c)
%d)
figure(4);
num_G_d= 1;
den_G_d=(s+2)^2*(s+5);
G_d=num_G_d/den_G_d;
823
nyquist(G_d)
%e)
figure(5);
num_G_e= 1*(s+5)*(s+1);
den_G_e=(s+50)*(s+2)^3;
G_e=num_G_e/den_G_e;
nyquist(G_e)
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
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 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
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 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.
or
if K < 0, the Nyquist cross the real axis at .So, for stability,
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
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.
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
833
K+1>0 K>-1
which is consistent with the results of the Nyquist diagrams.
8-21)
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
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
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)
1
=
( 2500 6 ) + j 70.6 = ( 2500 6 ) + j 70.6 ( 0.706 + j 0.06 )
2 2 2 3
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,
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 )
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
838
824(a) Kt = 0:
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
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
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
Forstability,
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
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -80
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
Frequency (rad/sec) Frequency (rad/sec)
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
MATLAB code:
s = tf('s')
%a)
figure(1);
K=1
num_GH_a= K*(s+1);
den_GH_a=(s-1)^2;
GH_a=num_GH_a/den_GH_a;
nyquist(GH_a)
%b)
figure(2);
K=1
num_GH_b= K*(s-1);
den_GH_b=(s+1)^2;
GH_b=num_GH_b/den_GH_b;
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).
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
2 10
1 5
0
0
-5
-1
0 -180
-50 -225
180
-270
0
-315
-180 -360
-2 -1 0 1 2 0
10 10 10 10 10 10
Frequency (rad/sec) Frequency (rad/sec)
846
Nyquist Diagram
2
1.5
-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):
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -20
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
Frequency (rad/sec) Frequency (rad/sec)
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
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 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.
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,
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);
GH_a=num_GH_a/den_GH_a;
nyquist(GH_a)
%b)
figure(2);
K=20
num_GH_b= K;
den_GH_b=s*(s+1)*(s+1);
GH_b=num_GH_b/den_GH_b;
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:
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -50
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
Frequency (rad/sec) Frequency (rad/sec)
852
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -70
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
Frequency (rad/sec) Frequency (rad/sec)
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
853
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -40
-90
0
-50 -135
-100 -180
0
-90
-180 -225
-270 P.M.: -47.8 deg
-360 Freq: 2.59 rad/sec
-450 -270
-1 0 1 2 0
10 10 10 10 10
Frequency (rad/sec) Frequency (rad/sec)
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
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);
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
OL_a = G_a*H_a
nyquist(OL_a)
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);
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
%CL_a = G_a/(1 + G_a*H_a);
OL_a = G_a*H_a
nyquist(OL_a)
859
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);
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
%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);
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
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
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
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
863
(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
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
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)
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
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
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
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
10 -20
5 -40
-60
0
-80
-5
-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
Frequency (rad/sec) Frequency (rad/sec)
K=59.9
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
10 40
5 20
0
0
-20
-5
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
Frequency (rad/sec) Frequency (rad/sec)
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);
G_a=num_G_a/den_G_a;
H_a = num_H_a/den_H_a;
CL_a = G_a/(1 + G_a*H_a);
margin(CL_a)
sisotool
atK=1,marginofstabilityisobservedasPM~=0:
K=1
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -80
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
Frequency (rad/sec) Frequency (rad/sec)
ThesystemisstableforK>1asfollows:K=0.6
882
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
1 0
0.5 -20
-40
0
-60
-0.5
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
Frequency (rad/sec) Frequency (rad/sec)
AndthesystemisunstableforK<1:K=3
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
1 20
0.5 0
-20
0
-40
-0.5
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
Frequency (rad/sec) Frequency (rad/sec)
*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)
TheBodeplotisdonewithK=1.
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)
TheBodeplotisdonewithK=1.
GM=50.21dB
PM=infinity.
ForGM=20dBKcanbe
increasedby30.21dBorK=32.4
ForPM=45deg.Kcanbe
increasedby38.24dB,or
K=81.66
(e)Seethebottomplot
TheBodeplotisdonewithK=1.
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
)
TheBodeplotisdonewithK=1.
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.
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.
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.
(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.
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);
GH_a=num_GH_a/den_GH_a;
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
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)
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
Frequency (rad/sec): 4.98
0
-0.5
-1
-1.5
-2 -1.5 -1 -0.5 0 0.5 1
Real Axis
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
Phasecrossoverfrequency=20.31rad/sec GM=15.11dB
(c)Gaincrossoverfrequency=19.1rad/sec PM=4.07deg
Phasecrossoverfrequency=20.31rad/sec GM=1.13dB
(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
894
843(a)Thegainisincreasedtofourtimesitsnominalvalue.Themagnitudecurveisraisedby12.04dB.
Gaincrossoverfrequency=10rad/sec PM=46deg
Phasecrossoverfrequency=20.31rad/sec GM=9.09dB
(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
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);
G_a=num_G_a/den_G_a;
CL_a = G_a/(1+G_a)
OL_a = G_a*1;
%(a)
figure(1)
nyquist(OL_a)
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
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
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)
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
Frequency (rad/sec): 33.5
-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.
Phasecrossoverfrequency=11.81rad/sec GM=10.51dB
8109
849(c)ClosedloopFrequencyResponse:
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
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
-50 Frequency (rad/sec): 0.712
Magnitude (dB): -3.05
-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)
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)
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
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
858(a)ForwardpathTransferFunction:
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
gaincrossoverfrequency=0rad/sec PM=infinite
(c)
1 s
G(s) =
(1 + s )(1 + 10 s )(1 + 2 s )
FromtheBodediagram,phasecrossoverfrequency=0.26rad/sec GM=25.44dB
gaincrossoverfrequency=0rad/sec PM=infinite
8121
858(continued)Bodediagramsforallthreeparts.
8122
859(a)ForwardpathTransferFunction:
s
e
G( s) =
(1 + 10 s )(1 + 25s )
FromtheBodediagram,phasecrossoverfrequency=0.37rad/sec GM=31.08dB
gaincrossoverfrequency=0rad/sec PM=infinite
(b)
1
G( s) =
(1 + 10s )(1 + 25s ) (1 + s + 0.5s 2
)
FromtheBodediagram,phasecrossoverfrequency=0.367rad/secGM=30.72dB
gaincrossoverfrequency=0rad/sec PM=infinite
(c)
(1 0.5s )
G( s) =
(1 + 10 s )(1 + 25s )(1 + 0.5 s )
FromtheBodediagram,phasecrossoverfrequency=0.3731rad/secGM=31.18dB
gaincrossoverfrequency=0rad/sec PM=infinite
8123
Plots859(ac)
8124
860SensitivityPlot:
8125
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)
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
Frequency (rad/sec) Frequency (rad/sec)
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
Theforwardpathtransferfunctionis Thecontrollertransferfunctionis
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.
92ForwardpathTransferFunction:
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
Theforwardpathtransferfunctionis Thecontrollertransferfunctionis
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
3 n=1.14
=1.09
K=1.42
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
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
InthefollowingsnapshotofCompensatorEditor,Cgainor K p issetto1,andthezerolocationissetto2.7,
resultingin44.2[deg]phasemargin,presentedinthefollowingfigure.
95
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Frequency (rad/sec) Frequency (rad/sec)
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))
s = tf('s')
K=1.0370;
num_G_a= 0.5*K;
den_G_a=s*(0.25*s^2+0.375*s+1);
G_a=num_G_a/den_G_a;
CL_a = G_a/(1+G_a)
BW = bandwidth(CL_a)
bode(CL_a);
sisotool;
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
90
Thus KD = = 0.09
1000
910
131.4
2 n = 10 + 1000 K D = 2 0.707 100 = 141.4 Thus KD = = 0.1314
1000
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)
Thephasemarginincreasesandthemaximumovershootdecreasesmonotonicallyas K D increases.
911
Zero/pole/gain:
1000(s+100)
(s+111.3)(s+898.7)
Gm=
912
UsethecursortoobtainthePOandtrvalues.
KDincreaseresultsintheminimumovershoot.
913
914
9-7)
PDcontrollerdesign
Theopenlooptransferfunctionofasystemis:
1
G(s) H (s) =
(2 s + 1)( s + 1)(0.5s + 1)
Thesolutionisverysimilarto94.Thetransferfunctionsareinsertedintosisotool,whereanotherrealzerois
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.
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -40
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
Frequency (rad/sec) Frequency (rad/sec)
PreliminaryMATLABcodefor97:
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
60
G ( s) H ( s) =
s (0.4 s + 1)( s + 1)( s + 6)
(i) Kv =10
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.
sisotoolwindow:
K p = 1 fixedto1,andzerolocationwaschangedintheentirerealaxis.However,2oftheclosedlooppoles
remainedintherighthandsideofSplaneintherootlocusdiagram,indicatinginstabilityforallKdvalues.
Solutionfor K v = 10 withPDcontrollerandPM=45[deg]doesnotexist.Unstablecloselooppolesare
indicatedintherootlocusdiagramofthefollowingfigure:
918
98)
PreliminaryMATLABcodefor98:
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
9-9)(a)ForwardpathTransferFunction:
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:
Select K D 0.0017
(b)BWmustbelessthan850rad/sec.
920
KD GM PM(deg) Mr BW(rad/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(%)
0.05 2.721 0
0.06 3.039 0
0.10 4.317 0
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)
ForwardpathTransferFunction:
200 ( 0.05 + K D s )
G ( s ) =
s ( s + 1)( s + 10)
AttributesofFrequencyResponse:
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.
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
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).
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
axis.Thissmallanglemeanshigherdampingofthepolesas=ArcCos(polesanglewithrealaxis).
928
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -250
-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
Frequency (rad/sec) Frequency (rad/sec)
929
Inthisparticularcase,closedloopcomplexpolescanbeobservedintheshownrootlocusdiagramatabout
0.81.2j.Thiscorrespondstodampingofabout:
cos(atan(1.2/0.8))=0.554 0.55
PreliminaryMATLABcodefor913:
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
931
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Frequency (rad/sec) Frequency (rad/sec)
9-15)
Leadcompensatorcontrollerdesign
Thebodediagramofthesystemwithoutleadcompensatorisshownbelow:
932
Bode Diagram
Gm = Inf dB (at Inf rad/sec) , Pm = 32.7 deg (at 17.3 rad/sec)
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
then from trial and error we found = 0.014088 and required gain would be 9.7185
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 /
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
-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.
(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.
Thequadraticportionofthecharacteristicequationis
2
(
s + 2 s + + 225 = 0
2
)
937
Thequadraticequationmustsatisfythecharacteristicequation.Usinglongdivisionandsolveforzero
remaindercondition.
s + (10 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)
2
125 + 20 3
KP = = 15765
.
100
(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)
(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
10 (1 + 0.1K P s )
G(s) =
(
s 1 + 0.1s + 0.01s
2
)
AttributesoftheFrequencyResponse:
940
When K P = 11
. and K I = 10, K v = 10, thephasemarginis31.51deg.,andismaximum.
Thecorrespondingrootsofthecharacteristicequationrootsare:
ReferringtheserootstotherootcontoursinProblem108(c),thecomplexrootscorrespondsto
arelativedampingratiothatisnearoptimal.
(b)SensitivityFunction:
941
Inthepresentcase,thesystemwiththePIcontrollerhasahighermaximumvalueforthesensitivity
function.
942
9-20)
(a)ForwardpathTransferFunction:
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)
(a)ForwardpathTransferFunction:
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.
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
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
Gm = 22.4 dB (at 44.9 rad/sec) , Pm = 52.5 deg (at 6.92 rad/sec)
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)
Correspondingly, the Bode diagram of closed loop system can be shown as:
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
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
was obtained as seen in the bode diagram of compensated loop:
Double Lead filter design:
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
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
T=1/alpha^0.5/CRover
lead=(1+T*alpha*s)/(1+T*s)
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;
Correspondingly, the Bode diagram of closed loop system can be shown as:
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)
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
was obtained as seen in the bode diagram of compensated loop:
Double Lead filter design:
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
Gm = 20.2 dB (at 20.6 rad/sec) , Pm = 41.9 deg (at 4.35 rad/sec)
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;
num_GH= 210*(Kp+Ki/s);
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))
T=1/alpha^0.5/CRover
lead=(1+T*alpha*s)/(1+T*s)
LT=GH*lead*lead
CL = LT/(1+LT);
953
figure(1)
Margin(GH)
figure(2)
Margin(LT)
figure(3)
Margin(CL)
Correspondingly, the Bode diagram of closed loop system can be shown as:
Bode Diagram
Gm = 19.3 dB (at 20.6 rad/sec) , Pm = 78.6 deg (at 5.95 rad/sec)
50
0
Magnitude (dB)
System: CL
Frequency (rad/sec): 7.44
-50
Magnitude (dB): -3.08
-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
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)
955
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -300
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
Frequency (rad/sec) Frequency (rad/sec)
Considering the slow dynamic of the ship, and the RHS zero, the crossover frequency is relatively low.
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
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
300 50
200
0
100
0
-50
-100
-200
-100
G.M.: 5.26 dB
-300 Freq: 81.6 rad/sec
-300 -200 -100 0 100 200
Stable loop
Bode Editor for Closed Loop 1 (CL1) -150
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
Frequency (rad/sec) Frequency (rad/sec)
959
. K
b) K lim sG s G s K Lim s K 30.3 , then
.
K 30.3
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:
Root Locus Editor for Open Loop 1 (OL1) Open-Loop Bode Editor for Open Loop 1 (OL1)
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
Bode Editor for Closed Loop 1 (CL1) -200
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
Frequency (rad/sec) Frequency (rad/sec)
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
ForwardpathTransferFunction:
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
)
1000 (1 + K D1 s )( s + 10 )
G(s) =
(
s s + 10 s + 100
2
)
962
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)
964
965
966
967
9-29)
968
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
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
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
Gm = Inf dB (at Inf rad/sec) , Pm = 22.6 deg (at 12.3 rad/sec)
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 .
1 5
20 23.5
1 5
970
1 5
20 23.5 17.5 6
1 5
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
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
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
2
For = 1 and n = 1rad/sec,thesecondordertermofthecharacteristicequationis s + 2 s + 1.
Dividingthecharacteristicequationbytheseondorderterm.
s + ( 0.00667 K D 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 P
0.01334 K D + 3.00667 ) s + 0.00667 K I 0.00667 K D + 2
Forzeroremainder,
FromEq.(2),
FromEq.(1),
975
ForwardpathTransferFunction:
G(s) =
(
0.00667 399.85s + 348.93s + 100
2
)
(
s s + 0.00667
2
)
CharacteristicEquation:
Roots:1,1,0.667
UnitstepResponse.
Themaximumovershootis20%.
976
9-34) a) As , therefore, or
b)
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:
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
UnitstepResponse.
Maximumovershoot=20.8%
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
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:
10,000 K D s
Geq ( s ) = 3 2
s + 10 s + 10,000 K P
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
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
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
m ( s ) 10 K p 5.12
= 2 2
in ( s ) s + 12 s + 10 K p s + 3.2 s + 5.12
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.
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!
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
KP PM GM Mr BW
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
becomes
areobtainedforvariousvaluesof K D1rangingfrom0.05to0.4.
KD1 PM GM Mr BW
Weseethatforvaluesof K D1greaterthan0.2,thephasemarginnolongerincreases,butthe
bandwidthincreaseswiththeincreasein K D1.Thuswechoose
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
3.6 9 2
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
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
993
Sincethezeroat0.5isveryclosetothepoleat0.5355,thesystemcanbeapproximatedbyasecond
ordersystem,
Y (s) 373.48
=
R(s) ( s + 9.3 )( s + 40.17 )
Theunitstepresponseisshownbelow.Theattributesoftheresponseare:
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
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
BodePlot.
995
996
9-43)(a)ForwardpathTransferFunction:
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:
Theattributesoftheunitstepresponseofthecompensatedsystemare:
(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
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:
Theattributesoftheunitstepresponseare:
Risetimetr=0.02278sec Settlingtimets=0.02828secMaximumovershoot=3.3%
998
999
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
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
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&
9100
Xc ( 2n + K D ) s + (n2 + K P )
= 2
Y s + ( 2n + K D ) s + ( 2n + K D )
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
(%)
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
Thephasemarginisatamaximumof54.69degwhenT=0.0006.Theperformanceworsensifthevalueofaislessthan
1000.
9103
946(a)BodePlot.
Theattributesofthefrequencyresponseare:
(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)
Thephasemarginismaximumat48.27degwhenT=0.0006.
9104
Next,wesetT=0.0006andreduceafrom1000.Wecanshowthatthephasemarginisnotvery
sensitivetothevariationofawhenaisnear1000.Theoptimalvalueofaisaround980,andthe
correspondingphasemarginis48.34deg.
Witha=980andT=0.0006,theattributesoftheunitstepresponseare:
(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)
Reducingthevalueofbfrom1000reducesthephasemargin.Thus,themaximumphasemarginof
100.4degisobtainedwithb=1000andT2=0.0004.Thetransferfunctionofthetwostagephase
leadcontrolleris
9105
(c)UnitstepResponses.
9106
x(t)
F
L
M
Here is an alternative representation including friction (damping) . In this case the angle is
measured differently.
1)
2)
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 ) )
A( K p + KDs)
= 2
R (s + K D s + A ( K p B )
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];
den =[1 Kd A*(Kp-B)];
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
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 )
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
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 )
and0.475 j 0.471whichcorrespondtoarelativedampingratioof0.707,whentheforwardpathgain
is4.5(asagainst200).Thus,thevalueofaofthephaselagcontrollerischosentobe
4.5
a= = 0.0225 SelectT=1000whichisalargenumber.
200
Then
UnitstepResponse.
9115
Maximumovershoot=13.6 tr=3.238sec ts=18.86sec
BodePlot(withphaselagcontroller,a=0.0225,T=1000)
9116
(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
9118
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)
Whena=1000,thebestvalueofTforamaximumphasemarginis0.00035,andPM=43.3deg.
Asitturnsoutvaryingthevalueofafrom1000doesnotimprovethephasemargin.Thusthe
transferfunctionofthecontrolleris
(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
(%)
Asseenfromtheresults,whena=0.09andforT 2000,themaximumovershootisless
than1%andthesettlingtimeislessthan2.5sec.WechooseT=2000anda=0.09.
Thecorrespondingfrequencydomaincharacteristicsare:
(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
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:
Theattributesoftheunitstepresponseare:
Maximumovershoot=6%tr=2.437sec ts=11.11sec
Comparingwiththephaselagcontrollerdesignedinpart(a)whichhasa=0.09andT=2000,
thetimeresponseattributesare:
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
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)
(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
KP tr ts MaxOvershoot
17 0.01113 0.01515 0
16 0.01184 0.01515 0
15 0.01303 0.01768 0
9129
UnitstepResponse:
(c)FrequencydomainDesignofPIController(KI=14.56)
9130
KP PM GM Mr BW MaxOvershoot tr ts
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:
(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.
plottedbelow,alongwiththatoftheuncompensatedsystem.
p 2 n MaxOvershoot
(%)
9132
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):
9135
955(a)ProcessTransferFunction
500 ( s + 10 )
Gp ( s) =
(
s s + 10 s + 1000
2
)
TheBodeplotisconstructedbelow.Thefrequencydomainattributesoftheuncompensated
systemare:
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
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:
BodePlots
9137
StepResponses
9138
(c)TimedomaindesignoftheNotchController
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
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
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 )
CharacteristicEquation:
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
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
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
ThusalltherequiredspecificationsstaywithintherequiredtoleranceswhenthevalueofKvariesby
plusandminus20%.
UnitstepResponses
9141
957Letthetransferfunctionofthecontrollerbe
Gc ( s ) =
(
200 s + 10 s + 50
2
)
( s + 100 ) 2
Theforwardpathtransferfunctionbecomes
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
4 3 5 2 6 7
a=8: s + 208 s + 2.116 10 s + 2.08 10 s + 10 = 0
4 3 5 6 7
a=12: s + 212 s + 2.124 10 + 2.12 10 s + 10 = 0
9142
UnitstepResponses:Allthreeresponsesfora=8,a=10,and12aresimilar.
958ForwardpathTransferFunction:
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
K ( s + 1)
Forrootloci, Geq ( s ) =
s
2
( s + 11)
9143
RootLocusPlot(Kvaries)
Therootlocishowthatarelativedampingratioof0.707canberealizedbytwovaluesofK.K=22
and59.3.Asstipulatedbytheproblem,weselectK=59.3.
9144
959ForwardpathTransferFunction:
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
3 2
s + 11s + 10.046 s + 4.6 = 0
960ForwardpathTransferFunction:
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:
TheunitstepresponsesofthesystemforK=20and44.6areshownbelow.
UnitstepResponses:
9146
961ForwardpathTransferFunction:
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
Theforwardpathtransferfunctionbecomes
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)
Theforwardpathtransferfunctionbecomes
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
(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
TheBodediagramisshownonthefollowingpage.Thefollowingfrequencyresponseattributes
areobtained:
BodePlot[parts(b),(c),and(d)]
9151
9152
963(a)TwostagePhaseleadControllerDesign.
Theuncompensatedsystemisunstable.PM=43.25degandGM=18.66dB.
Withasinglestagephaseleadcontroller,themaximumphasemarginthatcanberealizedaffectively
is12degrees.SettingthedesiredPMat11deg,wehavetheparametersofthesinglestagephase
1 + aT1 s 1 + 0.6057 s
is Gc1 ( s ) = =
1 + T1 s 1 + 0.00472 s
Startingwiththesinglestagecontrollercompensatedsystem,thesecondstageofthephaselead
controllerisdesignedtorealizeaphasemarginof60degrees.Theparametersofthesecondstage
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
ForwardpathTransferFunction:
Attributesofthefrequencyresponseofthecompensatedsystemare:
Theunitstepresponseisplottedbelow.Thetimeresponseattributesare:
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:
Theunitstepresponsehasalongrisetimeandsettlingtime.Theattributesare:
(c)LeadlagControllerDesign.
Fortheleadlagcontroller,wefirstdesignthephaselagportionfora40degreephasemargin.
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
Unitstepresponseattributes:Maximumovershoot=5.9% t s = 1512
. sec t r = 0.7882 sec
UnitstepResponses.
964(a)Theuncompensatedsystemhasthefollowingfrequencydomainattributes:
TheBodeplotof G p ( j ) showsthatthephasecurvedropsoffsharply,sothatthephaselead
controllerwouldnotbeveryeffective.Considerasinglestagephaselagcontroller.Thephase
9155
marginof60degreesisrealizedifthegaincrossoverismovedfrom2.8rad/secto0.8rad/sec.
Theattenuationofthephaselagcontrollerathighfrequenciesisapproximately15dB.
Choosinganattenuationof17.5dB,wecalculatethevalueofafrom
Theuppercornerfrequencyofthephaselagcontrollerischosentobeat1/aT=0.064rad/sec.
Thus,1/T=0.00854orT=117.13.Thetransferfunctionofthephaselagcontrolleris
1 + 15.63 s
Gc ( s ) =
1 + 117.13
Theforwardpathtransferfunctionis
FromtheBodeplotof G ( j ) ,thefollowingfrequencydomainattributesareobtained:
Theunitstepresponseattributesare:
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 )
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
9158
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:
Theattributesoftheunitstepresponseare:
BodePlots
9159
9160
965(c)TimedomainDesignofPIController.
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
Stateequations: Outputequation:
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
Stateequations: Outputequation:
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
Stateequations: Outputequation:
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
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
>> 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
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
-------------------------------------------------------------
-------------------------------------------------------------
Inputs:
A=| 0 1| B=|0|
C=|3 1| D=|0|
Dx = | 0 1|x + |0|u
107
y = |3 1|x + |0|u
Determinant of (s*I-A):
s +3s+2
s +3s+2
-1
-2
[ s+3 1 ]
[ ------------ ------------]
[ 2 2 ]
[s +3s+2 s + 3 s + 2]
[ ]
[ 2 s ]
[- ------------ ------------]
[ 2 2 ]
[ s +3s+2 s + 3 s + 2]
[ ]
s+3
------------
s +3s+2
[ 1 ]
[---------------]
[(s + 2) (s + 1)]
[ ]
[ s ]
[---------------]
[(s + 2) (s + 1)]
[ exp(-t) - exp(-2 t) ]
[ ]
Output Y(s):
s+3
---------------
(s + 2) (s + 1)
2 exp(-t) - exp(-2 t)
109
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))
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
Statetransitionmatrix:
Alternatively
1013
FromMATLABCommandWindow:
TheAmatrixis:
Amat=
01
21
1014
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
THERESTARESAMEASPARTA.
2
(b)Characteristicequation: ( s ) = sI A = s + 5s + 4 = 0 Eigenvalues: s = 4, 1
Statetransitionmatrix:
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
Statetransitionmatrix:
e 3t 0
(t ) = 3 t
0 e
Eigenvalues: s = 3, 3
2
(d)Characteristicequation: ( s ) = s 9 = 0
Statetransitionmatrix:
e3t 0
(t ) = 3 t
0 e
1015
Eigenvalues: s = j 2, j 2
2
(e)Characteristicequation: ( s ) = s + 4 = 0
Statetransitionmatrix:
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
Statetransitionmatrix:
e t 0 0
(t ) = 0
2 t 2 t
e te
0 0 e
2 t
e 5 t te
5 t
0
(t ) = 0
5 t 5 t
e te
0 0 e
5 t
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
(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).
1
1
0 1 0
[ (t ) ] = t
1
t
= = ( t )
1 e e 1 e
t t
e
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
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
(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
1022
TheAmatrixis:
Amat=
010
001
123
1023
CharacteristicPolynomial:
ans=
s^3+3*s^2+2*s+1
EigenvaluesofA=DiagonalCanonicalFormofAis:
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.
CharacteristicPolynomial:
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.
(2)Transferfunctionrelation:
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)
(3)Outputtransferfunction:
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.
(2)Transferfunctionrelation:
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
(3)Outputtransferfunction:
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)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
(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
(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
NOTE: the above order of calculations MUST BE followed in the order stated, otherwise you will get an
error.
1031
1032
Amat =
0 2 0
1 2 0
-1 0 1
Characteristic Polynomial:
ans =
s^3-3*s^2+2
Abar =
1.0000 0 0
0 2.7321 0
0 0 -0.7321
1033
T=
0 0.5591 0.8255
0 0.7637 -0.3022
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.
Characteristic Polynomial:
ans =
s^3-3*s^2+2
Transfer function:
2 s^2 - 3 s - 4
------------------------------
Zero/pole/gain:
2 (s-2.351) (s+0.8508)
--------------------------
1035
Smat =
0 2 4
1 2 6
1 1 -1
rankS =
Mmat =
0 -3 1
-3 1 0
1 0 0
1036
Ptran =
-2 2 0
0 -1 1
-4 -2 1
Abar =
0 1.0000 0.0000
0 -0.0000 1.0000
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:
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
1039
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):
Amat =
-1 0
0 -2
Characteristic Polynomial:
ans =
s^2+3*s+2
Abar =
-2 0
0 -1
T=
0 1
1 0
Characteristic Polynomial:
ans =
s^2+3*s+2
Transfer function:
-----
s+1
Zero/pole/gain:
-----
(s+1)
Smat =
2 -2
rankS =
Mmat =
3 1
1 0
Ptran =
4 2
0 0
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.
Amat =
-1 0
0 -2
Characteristic Polynomial:
ans =
s^2+3*s+2
1043
Abar =
-2 0
0 -1
T=
0 1
1 0
Characteristic Polynomial:
ans =
s^2+3*s+2
Transfer function:
Zero/pole/gain:
Vmat =
0 1
0 -2
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
Stateequations: Outputequation:
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
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
(1)Infiniteseriesexpansion:
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
(2)InverseLaplacetransform:
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
(1)Infiniteseriesexpansion:
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!
(2)InverseLaplacetransform:
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
(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
Foraunitstepfunctioninput, u s ( t ) = 1 / s.
1048
2
(c)Characteristicequation: ( s ) = s + 80.65 s + 322.58 = 0
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
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
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.
0 1
1049
AutomaticControlSystems,9thEdition
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))
0 2
From
m equation (1
1) and (2):
1
1026(a)Fo
orwardpathtrransferfunctio
on: C
Closedlooptr
ransferfunctio
on:
(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
'
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
Statetransitionmatrix:
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
Statetransitionmatrix:
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 )
0 1 0 0
A1 = 0 0 1 B1 = 0 C1 = [1 0 0]
3 2 1 1
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
Stateequations: x& ( t ) = Ax
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)
Stattediagram:
1063
AutomaticControlSystems,9thEdition
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)
Sttateequationss: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 0 0 1
A = 0 2 0 B = 1
0 0 10 1
(d)
1064
AutomaticControlSystems,9thEdition
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:
Staateequations: x& ( t ) = Ax
A ( t ) + Bu ( t )
5 1 0 0
A = 0 2 1 B=0
0 0 1.5 10
(b)
1065
AutomaticControlSystems,9thEdition
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:
Stateequations: x& ( t ) = Ax
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:
Staateequations: x& ( t ) = Ax
A ( t ) + Bu ( t )
0 1 0 0
A = 0 10 1 B = 0
0 0 2 5
(d)
1066
AutomaticControlSystems,9thEdition
A Chaptter10Solutio
ons Golnarraghi,Kuo
Y (s) 1
G(s) = =
U (s) (
s ( s + 5) s + 2s + 2
2
)
S
Statediagram:
Staateequations: x& ( t ) = Ax
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
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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:
1
1042(a)
Stattediagrambydirectdecompposition:
1072
AutomaticControlSystems,9thEdition
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
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 )
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
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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
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
Sissingular.Thesystemisunobservable.
(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
Sissingular.Thesystemisunobservable.
1079
AutomaticControlSystems,9thEdition
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
AutomaticControlSystems,9thEdition
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.
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
CharacteristicPolynomial:
ans=
s^2+1613/20*s+16129/50
EigenvaluesofA=DiagonalCanonicalFormofAis:
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
Continuoustimemodel.
CharacteristicPolynomial:
ans=
s^2+1613/20*s+16129/50
EquivalentTransferFunctionModelis:
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 :
The A B K matrixoftheclosedloopsystem
1.5028E+023.1938E+016.2112E+007.4534E+00
TheBvector
0.0000E+00
7.3200E02
0.0000E+00
1088
9.7600E02
TimeResponses:
1060(b) Thefeedbackgains,from k1 to k2 :
The A B K matrixoftheclosedloopsystem
7.0051E+021.2350E+029.9379E+015.9627E+01
TheBvector
0.0000E+00
7.3200E02
0.0000E+00
9.7600E02
TimeResponses:
1061ThesolutionsusingMATLAB
1090
(a) Thefeedbackgains,from k1 to k2 :
The A B K matrixoftheclosedloopsystem
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
CharacteristicEquation:
1093
s + 2 + k1 1
sI A = = s + ( 2 + k1 ) s + k 2 = 0
2
k2 s
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
10-64)
The equations of motion from Problem 4-21 are obtained (by ignoring all the pendulum inertia term):
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
CharacteristicPolynomial:
ans=
s^449/6*s^2
EigenvaluesofA=DiagonalCanonicalFormofAis:
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
Continuoustimemodel.
CharacteristicPolynomial:
ans=
s^449/6*s^2
EquivalentTransferFunctionModelis:
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
AbarK=
01.000000
001.00000
0001.0000
k1k28.1667k3k4
SystemCharacteristicequationis:
k4*s^4+(8.1667k3)*s^3k2*sk1=0
>> collect((s-210)*(s-210)*(s+20)*(s-12))
ans =
-10584000+s^4-412*s^3+40500*s^2+453600*s
10-65) If 3 and 0.707, then 1.414. The 2nd order desired characteristic equation of the
system is:
2 2 0 1
0 1
6 5
5 6 0 (2)
5 2
6 2
10103
10-66) Using ACSYS we can convert the system into transfer function form.
10104
10105
Amat =
-1 -2 -2
0 -1 1
1 0 -1
Characteristic Polynomial:
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
T=
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.
Characteristic Polynomial:
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
Zero/pole/gain:
3 (s+1.333) (s+1)
---------------------------------
Smat =
2 -4 0
0 1 0
1 1 -5
rankS =
Mmat =
5 3 1
3 1 0
1 0 0
Ptran =
-2 2 2
3 1 0
3 4 1
Abar =
0 1.0000 0
0 0 1.0000
10108
Bbar =
Cbar =
4 7 3
Dbar =
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
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
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.
If , , then
10110
or
2 1
Therefore:
2 1
1 1
0
0
2 1
10111
1 1
0
1
0
1 1
1
1 1
1
1
1
10112