Chapter 4 PDF

C H A P T E R 4
Feedback Control System

Characteristics
Exercises
)
E4.1 (a) The system sensitivity to τ is given by
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
SτT = SG
T G
an ing rnin tors igh
Sτ .
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
In this case, we have

of or stu e o tat
ity s w g us d S
is
te f t ss th nite
T 1 1 3s + 1
SG = = 100 = 3s + 101
e rt ss fo U
gr hi in e
th a a ly by
1 + GH(s)
k
1 + 3s+1
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
and
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
−τ s −3s
SτG =
a his
= ,
T
τs + 1 3s + 1
where τ = 3. Therefore,
−3s
SτT = .
3s + 101
(b) The closed-loop transfer function is
G(s) 100 100/101 0.99

T (s) = = = 3 = ,
1 + GH(s) 3s + 101 101 s + 1
τc s + 1
where the time-constant τc = 3/101 = 0.0297 second.

E4.2 (a) The system sensitivity to K2 is
T ∂T K2 1
SK = = .
2
∂K2 T 1 + K1 K2
134
© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
Exercises 135
(b) The transfer function from Td (s) to Vo (s) is
K2
Vo (s) = Td (s) .
1 + K1 K2
(c) We would select K1 ≫ 1, so that the transfer function from Td (s) to

Vo (s) is small.
E4.3 (a) The tracking error, E(s) = R(s) − Y (s), is given by
R(s) A/s
E(s) = = .
1 + G(s) 1 + K/(s + 10)2
The steady-state error (computed using the final value theorem) is

 
A A
ess = lim sE(s) = lim  = .
)
eb
K
er or in ing
1 + K/100
ed id n
s→0 s→0 1+
W
no the iss tea s
itt W tio
w
t p W em ch
(s+10)2
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
(b) A disturbance would be the wind shaking the robot arm.

e in nt ns co
D
th k ( de f i es
of or stu e o tat
E4.4 (a) The tracking error, E(s) = R(s) − Y (s), is given by

ity s w g us d S
is
te f t ss th nite
e rt ss fo U
R(s)
gr hi in e
th a a ly by
E(s) = .
in o e r
y y p d le d
ro n an o te
1 + KG(s)
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
The steady-state position error is computed (using the final value

o eir is rk
th nd wo
theorem) to be
a his
T
 
A/s A

ess = lim s = lim  =0.
s→0 1 + KG(s) s→0 1 + 100K
s(τ s+1)
(b) The ramp input of 0.1 m/sec is given by
0.1
R(s) = .
s2
Then, using the final value theorem, we have
  " #
0.1/s2  0.1
ess = lim s  100K
= lim ,
s→0 1 + s(τ s+1) s→0 s + 100K
τ s+1
or
0.1 0.001
ess = = .
100K K
136 CHAPTER 4 Feedback Control System Characteristics
We desire ess ≤ 0.0001 m, so

0.001
K≥ = 10 .
0.0001
E4.5 The closed-loop transfer function is
100K
T (s) = .
s2 + bs + 100K
The tracking error is
s(s + b) 1
E(s) = [1 − T (s)] R(s) = ,
s2 + bs + 100K s2
where we let R(s) = 1/s2 . Using the final value theorem we obtain the
steady-state tracking error as
b
)
eb
er or in ing
ess = lim sE(s) = .
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
s→0 100K
e
d on g. in t la
m ld a
r
If we require that b < 10K then the steady-state error is less than 0.1 to
or ud a uc y
w cl le tr p
e in nt ns co
the ramp input.

th k ( de f i es
of or stu e o tat
ity s w g us d S
E4.6 (a) The sensitivity is

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
" #
∂T p −s4 − 15s2 + 3s + 10
in o e r
y y p d le d
SpT
ro n an o te
= =p T (s).
st f a s d s ec
∂p T (s2 + ps + 10)2
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
(b) The tracking error is

a his
T
s3 + (2p − 1)s2 + (4 − p)s − 7 − p

E(s) = [1 − T (s)] R(s) = R(s)
s3 + 2ps2 + 4s + 3 − p
Using the final value theorem with R(s) = 1/s we obtain the steady-
state tracking error as
−7 − p
ess = lim sE(s) = .
s→0 3−p
E4.7 The light bounces off the surface of the slide and into a detector. If the
light fails to hit the detector dead center, the unbalanced electric signal
causes the motor to adjust the position of the light source, and simulta-
neously the lens.
0.1s + 1
T (s) = .
s2 + 1.1s + 1
The step response is shown in Figure E4.8.
Exercises 137
1.4
1.2
0.8
y(t)/A
0.6
0.4
0.2
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
0
r
0 2 4 6 8 10 12
or ud a uc y
w cl le tr p
e in nt ns co
Time (seconds)
th k ( de f i es
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
FIGURE E4.8
e rt ss fo U
gr hi in e
Step response.
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
(a) The closed-loop transfer function is

ill le u vi pr
E4.9
w r sa co pro is
o eir is rk
th nd wo
KK1
a his
T (s) = .
T
s + K1 (K + K2 )
(b) The sensitivities are
T ∂T /T s + K1 K2
SK = =
∂K/K s + K1 (K + K2 )
and
T s
SK = .
1
s + K1 (K + K2 )
(c) The transfer function from Td (s) to Y (s) is
Y (s) −1
= .
Td (s) s + K1 (K2 + K)
Therefore, since E(s) = −KY (s) (when R(s) = 0), we have
K
E(s) = Td (s)
s + K1 (K2 + K)
and
K
ess = lim sE(s) = .
s→0 K1 (K + K2 )
(d) With K = K2 = 1, we have
K1
T (s) = .
s + 2K1
Then,
K1 1
Y (s) =
s + 2K1 s
and
1h i
y(t) = 1 − e−2K1 t u(t) ,
2
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
where u(t) is the unit step function. Therefore, select K1 = 10 for the
fastest response.
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es

of or stu e o tat
ity s w g us d S
is
te f t ss th nite
46.24K(s + 50)(s + 425)

e rt ss fo U
T (s) = .
gr hi in e
th a a ly by
(s + 200)(s + 425)(s2 + 16.7s + 72.9) + 19652K(s + 50)

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
The steady-state error is determined to be

ill le u vi pr
w r sa co pro is
o eir is rk
1
th nd wo
ess = lim sE(s) = lim s(1 − T (s))

a his
s→0 s→0 s
T
= 1 − lim T (s) = 1 − T (0)

s→0
6.3
= .
6.3 + K
The plots of the steady-state error versus K and the percent overshoot
P.O. versus K are shown in Figure E4.10 for
40 ≤ K ≤ 400.
Exercises 139
0.14 45
0.12
40
0.1
35
percent overshoot (%)

steady-state error
0.08
30
0.06
25
0.04
20
0.02
0 15
)
eb
0 100 200 300 400 0 100 200 300 400
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
K K
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
FIGURE E4.10
e in nt ns co
D
th k ( de f i es
(a) Steady-state error. (b) Percent overshoot.

of or stu e o tat
ity s w g us d S
is
te f t ss th nite
(a) The closed-loop transfer function is

e rt ss fo U
E4.11
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
K(s2 + 5s + 6)
st f a s d s ec
G(s)
de o rse de ot
T (s) = = 3
ill le u vi pr
1 + G(s)H(s) s + 15s2 + 56s + 60 + 14K

w r sa co pro is
o eir is rk
th nd wo
(b) With E(s) = R(s) − Y (s) we obtain

a his
T
G(s) 1 − G(s)(1 − H(s))

E(s) = 1 − R(s) = R(s)
1 + G(s)H(s) 1 + G(s)H(s)
s3 + (15 − K)s2 + (56 − 5K)s + (60 + 8K) 1
= · .
s3 + 15s2 + 56s + 60 + 14K s
Then, using the final value theorem we find
(60 + 8K)
lim sE(s) = .
s→0 60 + 14K
(c) The transfer function from the disturbance Td (s) to the output is
1 s3 + 15s2 + 56s + 60
Y (s) = Td (s) = 3 Td (s) .
1 + G(s)H(s) s + 15s2 + 56s + 60 + 14K
The steady-state error to a unit step disturbance is
s3 + 15s2 + 56s + 60 1 60
lim sY (s) = lim s 3 2
· = .
s→0 s→0 s + 15s + 56s + 60 + 14K s 60 + 14K
(d) The sensitivity is
T ∂T K ∂T ∂G K
SK = =
∂K T ∂G ∂K T
1 K 1 + G(s)H(s) 1

= 2
= .
(1 + G(s)H(s)) s + 10 G(s) 1 + G(s)H(s)
E4.12 (a) The closed-loop transfer function is
Gc (s)G(s) 100K1 (s + 5)
T (s) = = 2 .
1 + Gc (s)G(s)H(s) s + 105s + (500 + 100K1 K2 )
The steady-state tracking error is
1 − Gc (s)G(s)(1 − H(s))

E(s) = R(s) − Y (s) = R(s)
1 + Gc (s)G(s)H(s)
s2 + (105 − 100K1 )s + 500 − 100K1 (5 − K2 ) 1
)
eb
er or in ing
ed id n
W
no the iss tea s
= ·
itt W tio
w
t p W em ch
e
s2 + 105s + 500 + 100K1 K2
d on g. in t la
m ld a
s
r
or ud a uc y
w cl le tr p
and
e in nt ns co
D
th k ( de f i es
of or stu e o tat
5 − K1 (5 − K2 )
ity s w g us d S
lim sE(s) = .
is
te f t ss th nite
s→0 5 + K1 K2
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
(b) The transfer function from the noise disturbance N (s) to the output
ro n an o te
st f a s d s ec
de o rse de ot
Y (s) is
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
−Gc (s)G(s)H(s) −100K1 K2

a his
Y (s) = N (s) = 2 N (s) .

T
1 + Gc (s)G(s)H(s) s + 105s + (500 + 100K1 K2 )
The steady-state error to a unit step N (s) = 1/s is

−100K1 K2 1 −K1 K2

lim sY (s) = lim s 2 · = .
s→0 s→0 s + 105s + (500 + 100K1 K2 ) s 5 + K1 K2
(c) The design trade-off would be to make K1 K2 as large as possible

to improve tracking performance while keeping K1 K2 as small as
possible to reject the noise.
K
T (s) = .
s2 + 50s + K
The sensitivity is
T ∂T /T s2 + 50s
SK = = 2 .
∂K/K s + 50s + K
Exercises 141

Gc (s)G(s) K
T (s) = = 2 .
1 + Gc (s)G(s)H(s) s + K1 s + K
The sensitivity is
T ∂T /T sK1
SK = =− 2 .
1
∂K1 /K1 s + K1 s + K
(b) You would make K as large as possible to reduce the sensitivity to
changes in K1 . But the design trade-off would be to keep K as small
as possible to reject measurement noise.
Gc (s)G(s) 100
T (s) = = 2 .
1 + Gc (s)G(s) s + 50s + 100
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
The steady-state tracking error is
r
or ud a uc y
w cl le tr p
e in nt ns co
1

E(s) = R(s) − Y (s) = R(s)
th k ( de f i es
of or stu e o tat
1 + Gc (s)G(s)
ity s w g us d S
is
te f t ss th nite
s2 + 50s 1
e rt ss fo U
= 2 ·
gr hi in e
th a a ly by
s + 50s + 100 s
in o e r
y y p d le d
ro n an o te
st f a s d s ec
and
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
lim sE(s) = 0 .
th nd wo
a his
s→0
T
(b) The transfer function from the disturbance Td (s) to the output Y (s)
is
1

Y (s) = 2 Td (s) .
s + 50s + 100
The steady-state error to a unit step Td (s) = 1/s is
1 1 1

lim sY (s) = lim s 2
· = .
s→0 s→0 s + 50s + 100 s 100
Problems
P4.1 The tank level control block diagram is shown in Figure P4.1.
dH
+ +
DQ1 G 1(s) DH
- +
FIGURE P4.1
Tank level control block diagram.
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
(a) For the open-loop system the transfer function is

r
or ud a uc y
w cl le tr p
e in nt ns co
R
th k ( de f i es
of or stu e o tat
G1 (s) = .
ity s w g us d S
RCs + 1
is
te f t ss th nite
e rt ss fo U
Thus,
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
∂G1 R 1
de o rse de ot
G1
SR = · = .
ill le u vi pr
∂R G1 RCs + 1
w r sa co pro is
o eir is rk
th nd wo
For the closed-loop system, the transfer function is

a his
T
G1 R
T (s) = = .
1 + KG1 RCs + 1 + KR
Thus,
T ∂T R 1
SR = · = ,
∂R T RCs + 1 + KR
and
T ∂T K −KR
SK = · = .
∂K T RCs + 1 + KR
(b) For the open-loop system
∆H(s)
=1.
δH(s)
All disturbances show up directly in the output, thus the open-loop
system has no capability to reject disturbances. On the other hand,
Problems 143
for the closed-loop system we have

∆H(s) 1 RCs + 1
= = .
δH(s) 1 + KG1 (s) RCs + 1 + KR
By selecting K large, we reduce the effects of any disturbances. For
example, consider a step disturbance. The steady-state error due to
the disturbance is
−(RCs + 1) A −A

ess = lim s = .
s→0 RCs + 1 + KR s 1 + KR
As K gets larger, the steady-state error magnitude gets smaller, as
desired.
(c) Consider the step input
A
∆Q1 (s) = .
)
eb
er or in ing
ed id n
s
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
Then, for the open-loop system we have

r
or ud a uc y
w cl le tr p
e in nt ns co
A
th k ( de f i es
of or stu e o tat
ess = lim s (1 − G1 ) = (1 − R)A .

ity s w g us d S
s→0 s
is
te f t ss th nite
e rt ss fo U
The steady-state error is zero when R = 1, but is sensitive to changes

gr hi in e
th a a ly by
k
in o e r
y y p d le d
in R. For the closed-loop system we have

ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
1 A A

w r sa co pro is
ess = lim s = .
o eir is rk
th nd wo
s→0 1 + KG1 s 1 + KR
a his
T
By selecting K large, the effect of the disturbance is reduced and is

relatively insensitive to changes in R.
P4.2 (a) The open-loop transfer function is
T (s) = Ka G(s) .
T is undefined and
Therefore, SK 1
T
SK a
=1.
The closed-loop transfer function is

Ka G(s)
T (s) = .
1 + Ka K1 G(s)
Therefore,
T ∂T K1 −Ka K1 G(s)
SK = · =
1
∂K1 T 1 + Ka K1 G(s)
and
T 1
SK = .
a
1 + K1 Ka G(s)
(b) The tracking error, E(s) = θd (s) − θ(s) = −θ(s), since θd (s) = 0. The
transfer function from the wave disturbance to the output θ(s) is
G(s)
θ(s) = Td (s) .
1 + K1 Ka G(s)
Consider a step disturbance input for the open- and closed-loop sys-
tems. For the open-loop system, we have
A
ess = − lim sG(s) = −A .
s→0 s
Thus, we see that the open-loop system does not have the capability
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
to reduce the effect of disturbances. For the closed-loop system, we
t p W em ch
e
d on g. in t la
m ld a
have
r
or ud a uc y
w cl le tr p
e in nt ns co
−G(s) A −Aωn2

th k ( de f i es
of or stu e o tat
ess = lim s = .
ity s w g us d S
s→0 1 + K1 Ka G(s) s 1 + K1 Ka ωn2

is
te f t ss th nite
e rt ss fo U
gr hi in e
We see that the larger we make K1 Ka , that smaller the effect of the
th a a ly by
k
in o e r
y y p d le d
ro n an o te
wave disturbance on the output in steady-state.

st f a s d s ec
de o rse de ot
ill le u vi pr
P4.3 (a) The open-loop transfer function is

w r sa co pro is
o eir is rk
th nd wo
a his
K
T
G(s) =
τs + 1
where K = k1 ka Eb . Then, computing the sensitivity yields
G
SK =1.
The closed-loop system transfer function is

K
T (s) = .
τ s + KKth + 1
Similarly, computing the sensitivity yields
T 1 τs + 1
SK = = .
1 + Kth G(s) τ s + 1 + KKth
(b) For the closed-loop system
1/(τ s + 1) Te (s)
T (s) = Te (s) ≈
1 + KKth /(τ s + 1) KKth
Problems 145
when KKth ≫ 1. So, by choosing KKth large, we can reduce the

effect of the disturbance. This cannot be done with the open-loop
system.
(c) Consider the step input
A
Edes (s) = .
s
The tracking error for the open-loop system is
E(s) = Edes (s) − T (s) .
Thus,
K A

ess = lim s 1 − = (1 − K)A .
s→0 τs + 1 s
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
So, ess = 0 when K = 1, but is sensitive to changes in K. The tracking
error for the closed-loop system is

r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
τ s + 1 + K(Kth − 1)
ity s w g us d S
E(s) = Edes (s)

is
te f t ss th nite
τ s + 1 + KKth
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
and
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
τ s + 1 + K(Kth − 1) A A(1 + K(Kth − 1))

w r sa co pro is
ess = lim sE(s) = lim s = .

o eir is rk
th nd wo
s→0 s→0 τ s + 1 + KKth s 1 + KKth

a his
T
Selecting Kth = 1 and K ≫ 1 reduces the steady-state error.

P4.4 (a) The overall transfer function is
Y (s) M G(s) + U QG(s)

T (s) = = .
R(s) 1 + QG(s)
(b) From Eq. (4.16) in Dorf & Bishop, we have

T N D
SG = SG − SG .
In our case, we find that

N
SG =1,
and
D QG(s)
SG = .
1 + QG(s)
Thus,
T QG(s) 1
SG =1− = .
1 + QG(s) 1 + QG(s)
(c) The sensitivity does not depend upon U (s) or M (s).

P4.5 The closed-loop transfer function is
G1 G(s)
T (s) = .
1 + G1 G(s)
(a) The sensitivity of T (s) to changes in ka is

1
SkTa = .
1 + G1 G(s)
)
(b) The transfer function from Td (s) to θ(s) is
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
G(s)
θ(s) = Td (s) .
r
or ud a uc y
w cl le tr p
1 + G1 G(s)
e in nt ns co
D
th k ( de f i es
of or stu e o tat
ity s w g us d S
Since we want θ(s) due to a disturbance, E(s) = −θ(s) and

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
−G(s) 10 −1

k
in o e r
y y p d le d
ess = lim sE(s) = lim s = .

ro n an o te
1 + G1 G(s) s ka
st f a s d s ec
s→0 s→0
de o rse de ot
ill le u vi pr
w r sa co pro is
Since our maximum desired error magnitude is

o eir is rk
th nd wo
a his
0.10o · π
T
ess = = 0.001745 rad ,

180
we select
ka ≥ 573 .
(c) The open-loop transfer function is
θ(s) = G(s)Td (s) .
So,
10

ess = − lim sG(s) →∞.
s→0 s
G1 G(s)
T (s) = .
1 + G1 G(s)
Problems 147
(a) The sensitivity is
T 1 (τ1 s + 1)(τe s + 1)
SK = = .
e
1 + G1 G(s) (τ1 s + 1)(τe s + 1) + K1 Ke
(b) The speed is affected by the load torque through the transfer function
−Kg G(s)
V (s) = ∆Td (s) .
1 + GG1 (s)
(c) Let R(s) = 30/s , and Ke K1 ≫ 1 . When the car stalls, V (s) = 0.
Using the final value theorem, we find
−Kg G(s) ∆d G1 G(s) 30

lim s + lim s
s→0 1 + GG1 (s) s s→0 1 + GG1 (s) s
+Kg Ke K1 Ke

= −∆d + 30 .
1 + Ke K1 1 + K1 Ke
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
Since Ke K1 ≫ 1, we have
r
or ud a uc y
w cl le tr p
Kg

e in nt ns co
Vss = −∆d + 30 .
th k ( de f i es
of or stu e o tat
K1
ity s w g us d S
is
te f t ss th nite
When Vss = 0, we have

e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
30K1
ro n an o te
st f a s d s ec
∆d = .
de o rse de ot
Kg
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
Thus, if
a his
T
Kg
=2,
K1
then ∆d = 15 percent grade ( i.e. ∆d = 15 ft rise per 100 ft horizon-
tally) will stall the car.
P4.7 (a) Let
k2
G1 (s) = k1 , G2 (s) = , and H(s) = k3 + k4 s .
s(τ s + 1)
Then the transfer function from TL (s) to Y (s) is
G2 (s) k2
Y (s) = − TL (s) = − TL (s) .
1 + G1 G2 H(s) s(τ s + 1) + k1 k2 (k3 + k4 s)
(b) The sensitivity of the closed-loop system to k2 is
1
SkT2 = ,
1 + G1 G2 H(s)
where T (s) is the closed loop transfer function
G1 G2 (s) k1 k2
T (s) = = .
1 + G1 G2 H(s) s(τ s + 1) + k1 k2 (k3 + k4 s)
(c) The error is given by
E(s) = R(s) − T (s)R(s) .
With
1
R(s) = ,
s
we have
1 1
)
eb
ess = lim s(1 − T (s)) = 1 − T (0) = 1 − .
er or in ing
ed id n
W
no the iss tea s
itt W tio
s k3w
t p W em ch
s→0
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
P4.8 (a) The sensitivity is

e in nt ns co
D
th k ( de f i es
of or stu e o tat
ity s w g us d S
T 1 (0.1s + 1)(s2 + 25s + 200)

is
te f t ss th nite
SK = = .
(0.1s + 1)(s2 + 25s + 200) + 200K
e rt ss fo U
1 + Gc (s)G(s)
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
(b) The transfer function from Td (s) to Y (s) is

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
G(s) 200(0.1s + 1)
Y (s) = Td (s) = Td (s) .
a his
1 + Gc (s)G(s) (0.1s + 1)(s2 + 25s + 200) + 200K

T
P4.9 (a) Computing the derivative of R with respect to i yields
dR −0.201R
= .
di (i − 0.005)3/2
When vout = 35 volts, we have
35
i= = 7ma .
5000
At the operating point i = 7 ma, we find from Figure P4.9(b) in Dorf
& Bishop that R ≈ 20K (note: If we use the given formula, we find
that R ≈ 8.2K when i = 7 ma, thus we see that the formula is just
an approximation to the plot). Using R = 20K, we have
dR −0.402 × 104
= = −45 kohms/ma .
di 0.896 × 10−4
Problems 149
The transfer function (valid around the operating point) is
Vout (s) K
T (s) = =
Vin (s) (sτ + 1) + K 45I
5
K
= .
sτ + 1 + 9KI
The photosensor block diagram is shown in Figure P4.9.
K vout
vin ts + 1
+
-
)
45 I
eb
5
er or in ing
ed id n
W
no the iss tea s
itt W tio
i (ma)
w
t p W em ch
e
d on g. in t la
m ld a
r
or ud a uc y
FIGURE P4.9
w cl le tr p
e in nt ns co
Photosensor block diagram.

th k ( de f i es
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
(b) The sensitivity of the system to changes in K is

in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
τs + 1
ill le u vi pr
T
w r sa co pro is
SK = .
o eir is rk
τ s + 1 + 9KI
th nd wo
a his
T
P4.10 (a) and (b) The paper tension control block diagram is shown in Fig-
ure P4.10.
DT (s)
V 1 + DV 1(s)
k3
E o( s) - w ( s) -
Km o 1 T( s)
R(s ) 2 s
+ + ts + 1 +
-
2
k2 k1
Y (s)
FIGURE P4.10
Paper tension control block diagram.
(c) The closed-loop transfer function, Tc , is given by

2Km
T (s) s(τ s+1) 2Km
Tc (s) = = 4K k
= .
R(s) 1 + k1 s(τ s+1)
m 2
τ s + s + 4Kkm1 k2
2
The sensitivity of Tc to changes in Km is
Tc 1 s(τ s + 1)
SK = 4Km k2
= .
m
1+ k1 s(τ s+1) τ s2 + s + 4Kkm1 k2
(d) The transfer function from ∆V1 (s) to T (s) is

−1
s −1
T (s) = 4Km k2
∆V1 (s) = 4Km k2
∆V1 (s) .
1+ k1 s(τ s+1) s+ k1 (τ s+1)
When ∆V1 (s) = A/s, we have
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
−(τ s + 1) A
e
d on g. in t la
m ld a
T (s) = .
4Km k2 s
τ s2
+ s + k1
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
and
ity s w g us d S
is
te f t ss th nite
−Ak1
e rt ss fo U
lim T (t) = lim sT (s) = .

gr hi in e
th a a ly by
t→∞ s→0 4Km k2

in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
P4.11 (a) The closed-loop transfer function is

ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
Gc (s)G(s) K K
T (s) = = = .
a his
2
T
1 + Gc (s)G(s) (10s + 1)(2s + 1) + K 20s + 12s + 1 + K

T is
(b) The sensitivity SK
T ∂T K 20s2 + 12s + 1
SK = · = .
∂K T 20s2 + 12s + 1 + K
(c) Define E(s) = R(s) − Y (s). Then
" #
R(s) 20s2 + 12s + 1
E(s) = = R(s) .
1 + Gc (s)G(s) 20s2 + 12s + 1 + K
With
A
R(s) = ,
s
we have
A
ess = lim sE(s) = .
s→0 1+K
Problems 151
(d) We want |e(t)| ≤ 0.02A as t → ∞ . So,

A
0.02A ≥
K +1
implies
K ≥ 49 .
P4.12 (a) The two transfer functions are

K1 K2
T1 (s) =
s2 + 3s − 4 + 6K1 K2
and
K1 K2
T2 (s) = .
s2 + (3 − 2K1 + 2K2 )s − 4 + 8K2 + 2K1 − 4K1 K2
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
When K1 = K2 = 1, we find that
e
d on g. in t la
m ld a
r
or ud a uc y
1
w cl le tr p
e in nt ns co
T1 (s) = T2 (s) = .
th k ( de f i es
s2 + 3s + 2
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
T1
(b) The sensitivity SK is
e rt ss fo U
1
gr hi in e
th a a ly by
k
in o e r
y y p d le d
s2 + 3s − 4
ro n an o te
∂T1 K1 (s + 4)(s − 1)
st f a s d s ec
T1
SK = · = 2 = ,
de o rse de ot
ill le u vi pr
1
∂K1 T1 s + 3s − 4 + 6K1 K2 (s + 2)(s + 1)
w r sa co pro is
o eir is rk
th nd wo
T2
when K1 = K2 = 1. The sensitivity SK is
a his
1
T
T2 ∂T2 K1 s+4 s+4

SK = · = = ,
1
∂K1 T2 s + 4 − 2K1 s+2
when K1 = 1. Thus,
T1 T2 s−1
SK = SK .
1 1
s+1
P4.13 (a) Let N (s) = G1 (s) + kG2 (s) and Td (s) = G3 (s) + kG4 (s). Then
∂N k ∂D k G2 k G4 k
SkT = · − · = −
∂k N ∂k D G1 + kG2 G3 + kG4
k(G2 G3 − G1 G4 )
= .
(G1 + kG2 )(G3 + kG4 )
M G(s) + kU G(s) G1 (s) + kG2 (s)
T (s) = = .
1 + kGH(s) G3 (s) + kG4 (s)
Then using result from (a), we have

k(U G(s) − M G2 H(s))
SkT = .
(M G(s) + kU G(s))(1 + kGH(s))
G(s) 20(s + 2)
T (s) = = .
1 + G(s) s(s + a)(s + 1) + 20(s + 2)
Then
SaT = SaN − SaD ,
where N is the numerator and D is the denominator. We have
SaN = 0 .
)
eb
er or in ing
Let
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
p(s)
r
G(s) = ,
or ud a uc y
w cl le tr p
e in nt ns co
q(s)(s + a)
th k ( de f i es
of or stu e o tat
ity s w g us d S
where p(s) = 20(s + 2) and q(s) = s(s + 1). Then

is
te f t ss th nite
e rt ss fo U
gr hi in e
G(s) p(s)
th a a ly by
T (s) = = ,
in o e r
y y p d le d
ro n an o te
1 + G(s) q(s)(s + a) + p(s)

st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
and
o eir is rk
th nd wo
a his
dD a −aq(s) a 1
SaT = −SaD = −
T
· = =− · .
da D q(s)(s + a) + p(s) s + a 1 + G(s)
P4.15 (a) Let
1 G1 (s)
G1 (s) = and G2 (s) = .
(τ1 s + 1)(τ2 s + 1) 100
Then
G1 (s) G2 Gc (s)
T2 (s) = To (s) + T2d (s) .
1 + G2 Gc (s) 1 + G2 Gc (s)
(b) We can equivalently consider the case of a step input, T2d = A/s,
To = 0, and zero initial conditions. Thus,
Gc G2 5 A
T2 (s) = T2d = 2
,
1 + Gc G2 500s + 60s + 6 s
where Gc (s) = 500. The transient response is shown in the Fig-
ure P4.15 for a unit step input (A = 1).
Problems 153
Step Response
0.9 System: sys

Peak amplitude: 0.94
Overshoot (%): 12.8
0.8 At time (seconds): 34.3 System: sys
Time (seconds): 54.1
Amplitude: 0.846
0.7
0.6
Amplitude
0.5
0.4
0.3
0.2
0.1
0
0 10 20 30 40 50 60 70 80 90 100
)
Time (seconds)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
FIGURE P4.15
r
or ud a uc y
w cl le tr p
Two tank temperature control system response.

e in nt ns co
D
th k ( de f i es
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
(c) With
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
E(s) = T2d (s) − T2 (s) ,

ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
we have
w r sa co pro is
o eir is rk
th nd wo
a his
E(s) = T2d (s) − M (s)T2d (s)

T
where
G2 Gc (s)
M (s) = .
1 + G2 Gc (s)
Then
A 5 A
ess = lim s(1 − M (s)) = (1 − M (0))A = (1 − )A = .
s→0 s 6 6
P4.16 The closed-loop transfer function for the disturbance to the output is
Y (s) G(s)
= ,
Td (s) 1 + KG(s)
with R = 0. The steady-state deviation is
G(s) 1 G(0) 1

yss = lim s = = .
s→0 1 + KG(s) s 1 + KG(0) 1+K
So, with K = 8 we have yss = 1/9, and with K = 22 we have yss = 1/23.
P4.17 (a) The closed-loop transfer function is
θ(s) 600 3000
= 2
= 2 .
θd (s) 0.2s + s + 600 s + 5s + 3000
The solution for a step input is
θ(t) = 1 − e−2.5t (cos 54.71t + 0.0457 sin 54.71t)
(b) The transfer function from the disturbance to the output is

θ(s) −1
= 2
.
Td (s) 0.2s + s + 600
Thus, the steady-state step disturbance is
)
A
eb
er or in ing
ed id n
θss = − lim sθ(s) = .
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
600
m ld a
s→0
r
or ud a uc y
Therefore, the disturbance input magnitude reduced by 600 at the

w cl le tr p
e in nt ns co
D
th k ( de f i es
output.
of or stu e o tat
ity s w g us d S
(c) Using the final value theorem we have (for θd (s) = 1/s2 )
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
ess = lim sE(s) = lim s(1 − T (s))θd (s)

in o e r
y y p d le d
ro n an o te
s→0 s→0
st f a s d s ec
de o rse de ot
!
0.2s2 + s 1 1
ill le u vi pr
= lim s = .
w r sa co pro is
o eir is rk
s→0 0.2s2 + s + 600 s 2 600

th nd wo
a his
T
Advanced Problems 155
Advanced Problems
AP4.1 The plant transfer function is
R
Gp (s) = .
RCs + 1
The closed-loop output is given by
1 GGp (s)
H(s) = Q3 (s) + Hd (s) .
1 + GGp (s) 1 + GGp (s)
Therefore, with E(s) = Hd (s) − H(s), we have
−1
E(s) = Q3 (s) ,
1 + GGp (s)
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
since Hd (s) = 0.
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
e in nt ns co
(a) When G(s) = K, we have

th k ( de f i es
of or stu e o tat
ity s w g us d S
−1
is
te f t ss th nite
ess = lim sE(s) = .

e rt ss fo U
gr hi in e
1 + KR
th a a ly by
s→0
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
(b) When G(s) = K/s, we have

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
ess = lim sE(s) = 0 .

a his
s→0
T
AP4.2 Define
Km Gc (s)
G(s) = .
s(La s + Ra )(Js + f ) + Km Kb Gc (s)
Then,
nG(s)
θ(s)/θd (s) =
1 + nG(s)
and
1
E(s) = θd (s) .
1 + nG(s)
So,
1 A A AKb
ess = lim sE(s) = lim s = = .
s→0 s→0 1 + nG(s) s 1 + nG(0) Kb + n
When θd (s) = 0 and Td (s) = M/s, we have
n(La s + Ra )
θ(s)/Td (s) = .
s(La s + Ra )(Js + f ) + Km Kb + Km Gc n
If Gc (s) = K, then
−nM Ra
ess =
Km (Kb + nK)
and if Gc (s) = K/s, we determine that ess = 0.

AP4.3 (a) The input R(s) is
1 1
R(s) = −
s s2
)
eb
and the disturbance is Td (s) = 0. So,
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
1 1 − 1s
r
or ud a uc y
w cl le tr p
ess = lim s R(s) = lim = −0.8750 .

e in nt ns co
s→0 1 + G(s) s→0 1 + 10(2s+4)

th k ( de f i es
of or stu e o tat
7s(s+5)
ity s w g us d S
is
te f t ss th nite
(b) The error plot is shown in Figure AP4.3a.

e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
2
w r sa co pro is
o eir is rk
th nd wo
a his
T
1. 5
1
e(t)
0. 5
?-0.5
?-1
0 2 4 6 8 10
Time (sec)
FIGURE AP4.3
(a) Error plot with d(t)=0.
?-0.5
e(t) ?-1
?-1.5
-?2
0 2 4 6 8 10
)
eb
er or in ing
ed id n
Time (sec)
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
FIGURE AP4.3
r
or ud a uc y
w cl le tr p
CONTINUED: (b) Error plot with r(t)=0.

e in nt ns co
D
th k ( de f i es
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
(c) The transfer function from Td (s) to Y (s) (with R(s) = 0 ) is

e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
−70
ro n an o te
st f a s d s ec
Y (s)/Td (s) = .
de o rse de ot
7s2 + 55s + 40
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
The steady-state error due to a disturbance Td (s) = 1/s is

a his
T
−70 1
ess = lim s = −1.75 .
s→0 7s2 + 55s + 40 s
(d) The error e(t) is shown in Figure AP4.3b.
AP4.4 (a) The closed-loop transfer function is
Km K
ω(s)/V (s) = .
Ra Js2 + Kb Km s + Km KKt
With v(t) = t, we have V (s) = 1/s2 , and Td (s) = 0. Using the final
value theorem yields
Ra Js + Kb Km Kb 1
ess = lim sE(s) = lim = = .
s→0 s→0 Ra Js2 + Kb Km s + Km KKt KKt 12K
We desire that
1
ess = < 0.1 .
12K
Therefore, we should select K > 0.8333. For example, we can take

K = 2.
(b) The transfer function from Td (s) to ω(s) is given by
ω(s) −10s
= 2 .
Td (s) s + s + 24
The error plot is shown in Figure AP4.4, where E(s) = −Kt ω(s)
(V (s) = 0.)
Disturbance ramp response

0.25
0.2
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
0.15
Tracking error

r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
0.1
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
0.05
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
0
th nd wo
0 1 2 3 4 5 6 7 8 9 10
Time (s)
a his
T
FIGURE AP4.4
Error plot with a ramp disturbance input.
AP4.5 (a) The transfer function from the disturbance Td (s) to the output Y (s)
is
Y (s) −s
= 3 .
Td (s) s + 4s2 + 4s + K
The steady-state error (when Td (s) = 1/s) is
s 1
ess = lim s =0.
s→0 s3 + 4s2 + 4s + K s
Y (s) K
= 3 2
.
R(s) s + 4s + 4s + K
The steady-state error (when R(s) = 1/s2 ) is
1 s3 + 4s2 + 4s 4
ess = lim s(1 − T (s)) 2
= lim 3 2
= .
s→0 s s→0 s(s + 4s + 4s + K) K
(c) Let K = 8. Then,
Y (s) −s
= 3 .
Td (s) s + 4s2 + 4s + 8
The error plot is shown in Figure AP4.5, for r(t) = 0.
0.15
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
0.1
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
0.05
ity s w g us d S
is
te f t ss th nite
e(t)
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
0
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
w sa co ro is
or eir is p rk
th nd wo
-0.05
a is
Th
-0.1
0 2 4 6 8 10 12 14 16 18 20
Time (sec)
FIGURE AP4.5
Error plot with a step disturbance input and K =8.
AP4.6 (a) The transfer function is
Vo (s) 1 + RCs
= .
V (s) 2 + RCs
(b) The system sensitivity is defined as
∂G/G
SCG = .
∂C/C
Therefore, the sensitivity is determined to be

RCs 1
SCG = = .
(2 + RCs)(1 + RCs) 1+ 2
1+ 1
RCs RCs
(c) Let V (s) = 1/s. Then

1 + RCs 1 0.5 0.5RC
Vo (s) = = + .
2 + RCs s s RCs + 2
Taking the inverse Laplace transform yields
vo (t) = 0.5(1 + e−2t/RC )u(t)
where u(t) is the unit step function. A plot of vo (t) versus t/RC is
shown in Figure AP4.6.
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
r
1
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
0.95
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
0.9
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
0.85
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
0.8
w r sa co pro is
o eir is rk
th nd wo
Vo
0.75
a his
T
0.7
0.65
0.6
0.55
0.5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
t / RC
FIGURE AP4.6
Step response.
AP4.7 (a) The transfer function from Td (s) to Y (s) is
Y (s) s
=− .
Td (s) s(s + 1) + K
(b) The transfer function from N (s) to Y (s) is

Y (s) K
= .
N (s) s(s + 1) + K
(c) Let Td (s) = A/s and N (s) = B/s. Then,
s A K B
ess = −yss = lim s − lim s = −B .
s→0 s(s + 1) + K s s→0 s(s + 1) + K s
So, K has no effect on the steady-state errors. However, choosing
K = 100 will minimize the effects of the disturbance Td (s) during the
transient period.
AP4.8 (a) The closed-loop transfer function is
Kb
T (s) = .
s + Kb + 2
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
(b) The sensitivity is determined to be
r
or ud a uc y
w cl le tr p
∂T /T s+2
e in nt ns co
SbT = = .
th k ( de f i es
of or stu e o tat
∂b/b s + Kb + 2
ity s w g us d S
is
te f t ss th nite
(c) The transfer function from Td (s) to Y (s) is

e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
Y (s) b
st f a s d s ec
= .
de o rse de ot
ill le u vi pr
Td (s) s + Kb + 2
w r sa co pro is
o eir is rk
th nd wo
So, choose K as large as possible, to make Y (s)/Td (s) as “small” as

a his
T
possible. Thus, select
K = 50 .
This also minimizes SbT at low frequencies.
Design Problems
CDP4.1 The model of the traction drive, capstan roller, and linear slide was de-
veloped in CDP2.1:
θ(s) Km
= .
Va (s) s [(Lm s + Rm )(JT s + bm ) + Kb Km ]
The step response for the closed-loop system (with the tachometer not in
the loop) and various values of the controller gain Ka is shown below.
% System parameters
Ms=5.693; Mb=6.96; Jm=10.91e-03; r=31.75e-03;
bm=0.268; Km=0.8379; Kb=0.838; Rm=1.36; Lm=3.6e-03; Lm=0;
)
eb
er or in ing
% Controller gain
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
Ka=100;
% Motor and slide model

r
or ud a uc y
w cl le tr p
Jt=Jm+r^2*(Ms+Mb);
e in nt ns co
D
th k ( de f i es
num=[Km];
of or stu e o tat
den=[Lm*Jt Rm*Jt+Lm*bm Kb*Km+Rm*bm 0];

ity s w g us d S
is
sys=tf(num,den);
te f t ss th nite
e rt ss fo U
%Closed-loop tf and step response

gr hi in e
th a a ly by
sys_cl=feedback(Ka*sys,[1]);
in o e r
y y p d le d
ro n an o te
step(sys_cl)
st f a s d s ec
de o rse de ot
ill le u vi pr
1.5
w r sa co pro is
o eir is rk
th nd wo
a his
T
Ka=2
Ka=5
Ka=10
1
Ka=100
Theta step response
0.5
0
0 0.2 0.4 0.6 0.8 1
Time (sec)
Design Problems 163
DP4.1 (a) The transfer function from the load disturbance to the output speed
is
ω(s) −G(s) −s
= = 2 .
Td (s) 1 + Gc G(s) s + 4s + K
Thus, the effect on ω(s) (of a unit step disturbance) at steady-state
is
−s 1

lim ω(t) = lim s 2
=0.
t→∞ s→0 s + 4s + K s
We see that the load disturbance has no effect on the output at steady-
state.
(b) The system response for 10 ≤ K ≤ 25 is shown in Figure DP4.1.
K=10,12,16,18,20,23,25
)
eb
er or in ing
ed id n
100.04
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
100.02
r
or ud a uc y
w cl le tr p
K=25
e in nt ns co
100
th k ( de f i es
of or stu e o tat
99.98
ity s w g us d S
is
te f t ss th nite
99.96
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
w(t)
y y p d le d
99.94
ro n an o te
K=10
st f a s d s ec
de o rse de ot
ill le u vi pr
99.92
w sa o ro s i
or eir is p rk
th nd wo
99.9
a his
T
99.88
99.86
99.84
0 0.5 1 1.5 2 2.5 3
Time(sec)
FIGURE DP4.1
Speed control system response.
1
For example , if we select K = 16, then ωn = 4, ζ = 2, and the
response due to a unit step disturbance is
−s 1 −1

ω(s) = 2 = .
s + 4s + 16 s (s + 2)2 + 12
Hence, if we are originally at ω(t) = 100 for t < τ , we have
1 √
ω(t) = 100 − √ e−2t sin 12t t ≥ τ .
12
DP4.2 With θd = 0, we have
G(s) s
θ(s) = T (s) = 3
KK1 d 2
Td (s) .
1 + G(s) s s + 4s + 9s + KK1
For Td = A/s, we have

A
θ(s) = .
s3 + 4s2 + 9s + KK1
The system response to a unit step disturbance for various values of KK1
are shown in Figure DP4.2. From the plot we see that when KK1 is small
the response is slow but not oscillatory. On the other hand, when KK1
is large the response is fast but highly oscillatory. In fact, if KK1 > 35,
the system is unstable. Thus, we might select KK1 = 10 as a reasonable
trade-off between fast performance and stability.
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
Unit step response for KK1=1,5,10,15,20,25

r
or ud a uc y
0.12
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
0.1
ity s w g us d S
is
te f t ss th nite
KK1=1
e rt ss fo U
0.08
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
0.06
de o rse de ot
KK1=5
ill le u vi pr
w r sa co pro is
o eir is rk
0.04
th nd wo
a his
q
0.02
-0.02
-0.04 KK1=25
-0.06
0 1 2 3 4 5 6 7 8 9 10
time(sec)
FIGURE DP4.2
Aircraft roll angle control system response to a disturbance.
DP4.3 (a) The closed-loop transfer function is
ω(s) K
T (s) = = 2 .
ωd (s) s + 5s + KK1
Design Problems 165
Then,
s2 + 5s + K(K1 − 1) 1
E(s) = (1 − T (s))ωd (s) = .
s2 + 5s + KK1 s
So, if
0.99 < K1 < 1.01 ,
then
|ess | < 0.01 .
(b) The transfer function from Td (s) to ω(s) is

−s
ω(s) = Td (s) .
s2 + 5s + KK1
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
So, with E(s) = −ω(s) and Td (s) = 2/s2 , we have
r
or ud a uc y
w cl le tr p
e in nt ns co
2
th k ( de f i es
lim sE(s) = .
of or stu e o tat
KK1
ity s w g us d S
s→0
is
te f t ss th nite
e rt ss fo U
gr hi in e
Therefore, we select KK1 > 20 to obtain ess < 0.1.

th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
DP4.4 The steady-state error for a step input command is zero for any K1 . The
de o rse de ot
ill le u vi pr
transfer function from Td (s) to Y (s) is

w r sa co pro is
o eir is rk
th nd wo
a his
Y (s) G(s) 2
T
= = 3 2
.
Td (s) 1 + KG(s) s + 5s + 4s + 2K
Thus, the output at steady-state due to a step disturbance Td (s) = A/s

is
A
lim sY (s) = .
s→0 K
We want to maximize K to reduce the effect of the disturbance. As we will
see in Chapter 6, we cannot select K too high or the system will become
unstable. That is why the problem statement suggests a maximum gain
of K = 10. For the design we choose
K = 10 .
DP4.5 The transfer function from V (s) to Vo (s) is
ks
Vo (s)/V (s) =
s+a
where
R2 + R3 1
k= and a= .
R2 R1 C
Computing the step response, we find that
vo (t) = ke−at = 5e−100t .
Solving for R1 , R2 , R3 and C yields

R2
R1 C = 0.01 and =4.
R3
K/J
θ(s) = θd (s) .
)
eb
s2 + K/J
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
Since J > 0, the system is unstable when K < 0 and marginally

r
or ud a uc y
w cl le tr p
e in nt ns co
stable when K > 0.

th k ( de f i es
of or stu e o tat
(b) Since the system is marginally stable, the system response does not
ity s w g us d S
is
te f t ss th nite
have a steady-state value—it oscillates indefinitely.

e rt ss fo U
gr hi in e
th a a ly by
(c) The closed-loop transfer function is

in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
KD s + KP
w r sa co pro is
θ(s) = θd (s) .
o eir is rk
Js2 + KD s + KP
th nd wo
a his
T
The system is stable for all KD > 0 and KP > 0, given that J > 0.
(d) The tracking error E(s) = θd (s) − θ(s) is
Js2
E(s) = .
Js2 + KD s + KP
Therefore, using the final value theorem we obtain the steady-state
value
Js2 1
lim sE(s) = lim s 2
· =0.
s→0 s→0 Js + KD s + KP s
s 1
Y (s) = Td (s) = 2
s2 + Ks + 2K s + Ks + 2K
where the disturbance is a unit step Td (s) = 1/s. Considering the
poles of the closed-loop system, we find that when K > 8 the system
Design Problems 167
has two real poles. In that case the disturbance step response is
1
y(t) = √ e−αt − e−βt ,
K 2 − 8K
where
√ √
K− K 2 − 8K K+ K 2 − 8K
α= and β =
2 2
Bounding the maximum y(t) yields the inequality
1
|y(t)| = √
−αt
e − e−βt ≤ 0.05.

K 2 − 8K
We know that

−αt
e − e−βt ≤ 1,

)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
for any α and β computed as shown above where K > 8. So, if we
choose K such that

r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
1
of or stu e o tat
√ ≤ 0.05.
ity s w g us d S
K 2 − 8K
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
we will guarantee that the maximum bound of 0.05 is not exceeded.

k
in o e r
y y p d le d
ro n an o te
Solving for K yields

st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
K > 24.4.
o eir is rk
th nd wo
a his
T
For any K > 24.4 we know that the maximum value of the disturbance
step response will be less than 0.05. When K = 24.4 the maximum
unit step disturbance response is 0.035. Solving explicitly for K so
that the maximum is 0.05 we find that K = 16.3 (this was found
numerically since it is very difficult to obtain analytically).
(b) Since the system is type 2, we know that the steady-state value of the
disturbance step response is zero for a unit step disturbance.
DP4.8 (a) The sensitivities are
" #
∂T /T −s2 (τ2 s + 1) τ1
SτT1 = = T (s)
∂τ1 /τ1 K
and
" #
∂T /T −s2 (τ1 s + 1) τ2
SτT2 = = T (s)
∂τ2 /τ2 K
where we assume that K 6= 0.
(b) Computing the closed-loop transfer function yields

s (τ1 s + 1)
Y (s) = Td (s)
s (τ1 s + 1) (τ2 s + 1) + K
When Td (s) = 1/s, using the final value theorem we find that
s (τ1 s + 1)
lim sY (s) = lim =0
s→0 s→0 s (τ1 s + 1) (τ2 s + 1) + K
as long as K 6= 0. We assume here that final value theorem applies

(i.e., the system is stable, more on this in Chapter 6).
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his
T
Computer Problems 169
Computer Problems
CP4.1 The step response and an m-file script which generates the step response
is shown in Figure CP4.1. The closed-loop transfer function is
12
T (s) = .
s2 + 2s + 22
The percent overshoot is P.O. = 50.2% and the steady-state error is
ess = 0.45.
Step Response
0.9
0.8 System: sys_cl

Peak amplitude: 0.82
)
Overshoot (%): 50.2
eb
0.7
er or in ing
At time (sec): 0.67
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
0.6
r
or ud a uc y
num = [12]; den = [1 2 10];

w cl le tr p
Amplitude
e in nt ns co
sys = tf(num,den); 0.5

th k ( de f i es
of or stu e o tat
sys_cl = feedback(sys,[1])
ity s w g us d S
0.4
step(sys_cl)
is
te f t ss th nite
e rt ss fo U
gr hi in e
0.3
th a a ly by
k
in o e r
y y p d le d
ro n an so cte
0.2
e
de o rse de ot
ill le u vi pr
d
w r sa co pro is
0.1
o eir is rk
th nd wo
st a s
a his
0
T
0 1 2 3 4 5 6
Time (sec)
FIGURE CP4.1
Step response.
CP4.2 The transfer function is

4
G(s) = .
s2 + 2s + 20
An m-file script which generates the step response is shown in Figure CP4.2.
The step response is also shown in Figure CP4.2. The step response is
generated using the step function. In the script, the transfer function
numerator is represented by num and the denominator is represented by
den. The steady-state value is yss = 0.2 and the desired value is 1.0.
Therefore, the steady-state error is
ess = 0.8 .
Step Response
0.35
num=[4]; den=[1 2 20]; 0.3

sys = tf(num,den);
axis([0 6 0 1]);
0.25
t=[0:0.01:6];
step(sys,t) Amplitude
y = step(sys,t); 0.2
yss = y(length(t))
0.15
0.1
yss =
0.20 0.05
)
0
eb
0 1 2 3 4 5 6
er or in ing
ed id n
W
no the iss tea s
itt W tio
wTime (sec)
t p W em ch
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
e in nt ns co
FIGURE CP4.2
th k ( de f i es
Step response.
of or stu e o tat
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
CP4.3 The step responses and the m-file script which generates the step re-
in o e r
y y p d le d
ro n an o te
st f a s d s ec
sponses is shown in Figure CP4.3.

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his
T
7
K=10
K=200
6 K=500
K=[10,200,500]; 5
t=[0:0.01:7];
for i=1:3 4
num=5*K(i); den=[1 15 K(i)];
sys = tf(num,den)
3
y(:,i)= step(sys,t);
end
plot(t,y(:,1),t,y(:,2),'--',t,y(:,3),':') 2
legend('K=10','K=200','K=500',-1)
1
0
0 1 2 3 4 5 6 7
FIGURE CP4.3
Step responses for K = 10, 100, 500.
CP4.4 (a,b) The m-file and plots are shown in Figure CP4.4.
Step Response Step response

1.5
Amplitude
ng=1;dg=[1 2 0];sysg=tf(ng,dg); 1
K=10;
syscl=feedback(K*sysg,1); 0.5
figure(1)
subplot(211) 0
0 1 2 3 4 5 6
step(syscl) Time (seconds)
subplot(212)
Step Response Disturbance response
syst=feedback(sysg,K)
0.2
step(syst)
0.15
Amplitude
0.1
)
eb
0.05
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
0
0 1 2 3 4 5 6
Time (seconds)
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
FIGURE CP4.4
ity s w g us d S
Step response and disturbance response.

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
(c) The estimated steady-state tracking error due to a unit step input is
de o rse de ot
ill le u vi pr
zero, and the estimated steady-state tracking error to a unit distur-

w r sa co pro is
o eir is rk
bance is 0.1.
th nd wo
a his
(d) The estimated maximum tracking error due to a unit step input is
T
0.35, and the estimated maximum tracking error to a unit disturbance

is 0.14. The maximum occurs at approximately t = 1 s.
CP4.5 The step response and the m-file script which generates the step response
is shown in Figure CP4.5. The closed-loop transfer function is determined
to be
10
T (s) = .
s2 + 3.7s + 10
Using the m-file script, a trial-and-error search on k yields
k = 3.7 .
The percent overshoot P.O. = 10% and the steady-state value is 1, as

expected.
1.4
k = 3.7; % Final value of k=3.7 1.2

numcg = [10]; dencg = [1 k 0];
sys_o = tf(numcg,dencg);
sys_cl = feedback(sys_o,[1]) 1
t = [0:0.1:5];
[y,t] = step(sys_cl,t); 0.8
plot(t,y,[0 5],[1.1 1.1],'--'); grid
y(t)
xlabel('Time (sec)'); ylabel('y(t)');
0.6
0.4
Transfer function:
10 0.2
----------------
s^2 + 3.7 s + 10
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time (sec)
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
FIGURE CP4.5 w
t p W em ch
e
d on g. in t la
m ld a
Step response.
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
CP4.6 The closed-loop transfer function is

of or stu e o tat
ity s w g us d S
is
te f t ss th nite
K
T (s) =
e rt ss fo U
gr hi in e
th a a ly by
s−a+K
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
where K = 2. When a = 1 and R(s) = 1/s, the final value is

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
K
th nd wo
lim sT (s)R(s) = lim T (s) = =2.

a his
s→0 s→0 K −a
T
The output is within 2% of the final value at around t = 4.6 seconds. The
plot of the step responses for
a = 1, 0.5, 2, 5
is shown in Figure CP4.6. The output is unstable for
a > 2.
K=2; t=[0:0.1:5];
num=K*[1];
a=[1 0.5 2 5];
for i=1:4
den=[1 -a(i)]; sys = tf(num,den);
sys_cl = feedback(sys,[1]);
y(:,i)=step(sys_cl,t);
end
plot(t,y(:,1),t,y(:,2),':',t,y(:,3),'--',t,y(:,4),'-.')
axis([0 5 0 5]);
xlabel('Time (sec)'), ylabel('y(t)')
title('a=1 (solid); a=0.5 (dotted); a=2 (dashed); a=5 (dashdot)')
a=1 (solid); a=0.5 (dotted); a=2 (dashed); a=5 (dashdot)
5
4.5
)
3.5
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
3
r
or ud a uc y
w cl le tr p
y(t)
e in nt ns co
2.5
th k ( de f i es
of or stu e o tat
ity s w g us d S
2
is
te f t ss th nite
e rt ss fo U
1.5
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
1
de o rse de ot
ill le u vi pr
w r sa co pro is
0.5
o eir is rk
th nd wo
a his
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
T
Time (sec)
FIGURE CP4.6
Step response for a=1, 0.5, 2, and 5.
CP4.7 The transfer function from the disturbance to the output is

G(s) 1
T (s) = = 2
.
1 + K0 G(s) Js + bs + k + K0
The disturbance response is shown in Figure CP4.7. The compensated
system response is significantly reduced from the uncompensated system
response. The compensated system output is about 11 times less than the
uncompensated system output. So, closed-loop feedback has the advan-
tage of reducing the effect of unwanted disturbances on the output.
J=1; k=5; c=0.9;

num=[1/J]; den=[1 c/J k/J];
sys = tf(num,den);
t=[0:0.1:10];
%
yu=step(sys,t); % Part (a)
K0=50;
numk=[K0]; denk=[1]; sysk = tf(numk,denk);
sys_cl = feedback(sys,sysk);
yc=step(sys_cl,t); % Part (b)
plot(t,yu,t,yc,'--')
xlabel('Time (sec)'), ylabel('\theta')
title('Uncompensated response (solid) & Compensated response (dashed)')
Uncompensated response (solid) & Compensated response (dashed)
0.35
0.3
)
eb
er or in ing
ed id n
0.25
W
no the iss tea s
itt W tio
w
t p W em ch
e
d on g. in t la
m ld a
r
or ud a uc y
0.2
w cl le tr p
e in nt ns co
D
th k ( de f i es
of or stu e o tat
q
ity s w g us d S
0.15
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
0.1
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
0.05
o eir is rk
th nd wo
a his
T
0
0 1 2 3 4 5 6 7 8 9 10
Time (sec)
FIGURE CP4.7
Disturbance responses for both the uncompensated and compensated systems.
CP4.8 The step responses for the proportional and PI controller are shown in
Figure CP4.8. The steady-state tracking error for the proportional con-
troller is
ess = 0.33 .
Increasing the complexity of the controller from a proportional controller

to a proportional plus integral (PI) controller allows the closed-loop sys-
tem to track the unit step response with zero steady-state error. The cost
is controller complexity, which translates into higher costs ($).
numg=[10]; deng=[1 10]; sysg = tf(numg,deng);

t=[0:0.001:0.5];
% Part (a)
numc=[2]; denc=[1]; sysc = tf(numc,denc);
sys_o = series(sysc,sysg);
sys_cl = feedback(sys_o,[1]);
yk=step(sys_cl,t);
% Part (b)
numc=[2 20]; denc=[1 0]; sysc = tf(numc,denc);
sys_o = series(sysc,sysg);
sys_cl = feedback(sys_o,[1]);
yp=step(sys_cl,t);
%
plot(t,yk,t,yp,'--')
xlabel('Time (sec)'),ylabel('y(t)')
title('Proportional controller (solid) & PI controller (dashed)')
Proportional controller (solid) & PI controller (dashed)

1
)
eb
er or in ing
ed id n
W
no the iss tea s
itt W tio
w
t p W em ch
0.9
e
d on g. in t la
m ld a
r
or ud a uc y
w cl le tr p
0.8
e in nt ns co
D
th k ( de f i es
of or stu e o tat
0.7
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
0.6
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
y(t)
0.5
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
0.4
th nd wo
a his
T
0.3
0.2
0.1
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Time (sec)
FIGURE CP4.8
Step response for proportional controller and PI controller.
CP4.9 (a) The closed-loop transfer function is
G(s) 10s2 + 500s

T (s) = R(s) = 2 R(s) .
1 + G(s)H(s) s + 200s + 5000
The step response is shown in Figure CP4.9a.
(b) The response of the system to the sinusoidal disturbance

100
N (s) =
s2 + 100
is shown in Figure CP4.9b.
(c) In the steady-state, the magnitude of the peak response is 0.095 and
the frequency is 10 rad/sec (see Figure CP4.9b).
% Part (a)
ng=10*[1 0]; dg=[1 100]; sysg=tf(ng,dg); >>
nh=[5]; dh=[1 50]; sysh=tf(nh,dh); Transfer function:
sys=feedback(sysg,sysh) 10 s^2 + 500 s
figure(1) ------------------
step(sys) s^2 + 200 s + 5000
Step Response
% Part (b) 10
)
eb
sysn=-feedback(sysg*sysh,1)
er or in ing
ed id n
W
no the iss tea s
itt W tio
syss=tf([100],[1 0 100]); w
t p W em ch
8
e
d on g. in t la
m ld a
% This is the sinusoidal input
r
figure(2)
or ud a uc y
w cl le tr p
Amplitude
e in nt ns co
t=[0:0.001:7];
th k ( de f i es
of or stu e o tat
step(syss*sysn,t)
ity s w g us d S
4
is
te f t ss th nite
e rt ss fo U
gr hi in e
2
th a a ly by
k
in o e r
y y p d le d
ro n an o te
Step Response
st f a s d s ec
0.1
0
de o rse de ot
0 0.02 0.04 0.06 0.08 0.1 0.12

ill le u vi pr
0.08
w r sa co pro is
Time (sec)
o eir is rk
0.06
th nd wo
(a)
a his
0.04
T
0.02
Amplitude
−0.02
−0.04
−0.06
−0.08
−0.1
0 1 2 3 4 5 6 7
Time (sec)
(b)
FIGURE CP4.9
(a) Unit step response. (b) Response to sinusoidal noise input at ω = 10 rad/sec.
CP4.10 (a) The closed-loop transfer function is

Gc (s)G(s) K(s + 1)
T (s) = R(s) = 2
R(s) .
1 + G(s)Gc (s) (s + 15)(s + s + 6.5) + K(s + 1)
(b) The step responses are shown in Figure CP4.10a.
(c) The unit disturbance response of the system is shown in Figure CP4.10b.
The steady-state value is 0.14.
Step Response
0.7
K=5
0.2
K=10
0.6 K=50
0.18
0.16 System: syscl

0.5 Final Value: 0.14
0.14
Step response
0.4 0.12
Amplitude
0.1
0.3
0.08
0.2 0.06
0.04
0.1
0.02
)
eb
er or in ing
0 0
ed id n
W
no the iss tea s
itt W tio
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
w
t p W em ch
e
d on g. in t la
m ld a
Time (s) Time (sec)
(a) (b)
r
or ud a uc y
w cl le tr p
e in nt ns co
D
th k ( de f i es
FIGURE CP4.10
of or stu e o tat
(a) Unit step responses for K = [5, 10, 50]. (b) Disturbance unit step response.
ity s w g us d S
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by
k
in o e r
y y p d le d
ro n an o te
CP4.11 The m-file is shown in Figure CP4.11a and the step responses in Fig-
st f a s d s ec
de o rse de ot
ure CP4.11b.
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his
T
5
K=10
K=12
4 K=15
K=[10, 12, 15];

3
t=[0:0.1:20];
ng=[20]; dg=[1 4.5 64]; sysg=tf(ng,dg);
nh=[1]; dh=[1 1]; sysh=tf(nh,dh); 2
Step response
for i=1:length(K)
sys=K(i)*sysg; 1
syscl=feedback(sys,sysh)
y(:,i)= step(syscl,t); 0
end
plot(t,y(:,1),t,y(:,2),'--',t,y(:,3),':')
−1
xlabel('Time (s)')
ylabel('Step response')
legend('K=10','K=12','K=15',-1) −2
−3
0 5 10 15 20
Time (s)
(a) (b)
FIGURE CP4.11
(a) M-file script. (b) Unit step responses for K = [10, 12, 15].

Chapter 4 PDF

Uploaded by

Copyright:

Available Formats

Chapter 4 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chapter 4 PDF

Uploaded by

Copyright:

Available Formats

C H A P T E R 4

Feedback Control System

In this case, we have

G(s) 100 100/101 0.99

where the time-constant τc = 3/101 = 0.0297 second.

(b) The transfer function from Td (s) to Vo (s) is

(c) We would select K1 ≫ 1, so that the transfer function from Td (s) to

The steady-state error (computed using the final value theorem) is

(b) A disturbance would be the wind shaking the robot arm.

E4.4 (a) The tracking error, E(s) = R(s) − Y (s), is given by

The steady-state position error is computed (using the final value

(b) The ramp input of 0.1 m/sec is given by

We desire ess ≤ 0.0001 m, so

the ramp input.

E4.6 (a) The sensitivity is

(b) The tracking error is

s3 + (2p − 1)s2 + (4 − p)s − 7 − p

(a) The closed-loop transfer function is

E4.10 The closed-loop transfer function is

46.24K(s + 50)(s + 425)

(s + 200)(s + 425)(s2 + 16.7s + 72.9) + 19652K(s + 50)

The steady-state error is determined to be

ess = lim sE(s) = lim s(1 − T (s))

= 1 − lim T (s) = 1 − T (0)

percent overshoot (%)

(a) Steady-state error. (b) Percent overshoot.

(a) The closed-loop transfer function is

1 + G(s)H(s) s + 15s2 + 56s + 60 + 14K

(b) With E(s) = R(s) − Y (s) we obtain

G(s) 1 − G(s)(1 − H(s))

(d) The sensitivity is

E4.12 (a) The closed-loop transfer function is

−Gc (s)G(s)H(s) −100K1 K2

Y (s) = N (s) = 2 N (s) .

1 + Gc (s)G(s)H(s) s + 105s + (500 + 100K1 K2 )

The steady-state error to a unit step N (s) = 1/s is

(c) The design trade-off would be to make K1 K2 as large as possible

E4.14 (a) The closed-loop transfer function is

(a) For the open-loop system the transfer function is

For the closed-loop system, the transfer function is

for the closed-loop system we have

Then, for the open-loop system we have

ess = lim s (1 − G1 ) = (1 − R)A .

The steady-state error is zero when R = 1, but is sensitive to changes

in R. For the closed-loop system we have

By selecting K large, the effect of the disturbance is reduced and is

The closed-loop transfer function is

s→0 1 + K1 Ka G(s) s 1 + K1 Ka ωn2

wave disturbance on the output in steady-state.

P4.3 (a) The open-loop transfer function is

The closed-loop system transfer function is

(b) For the closed-loop system

when KKth ≫ 1. So, by choosing KKth large, we can reduce the

E(s) = Edes (s) − T (s) .

error for the closed-loop system is

E(s) = Edes (s)

τ s + 1 + K(Kth − 1) A A(1 + K(Kth − 1))

ess = lim sE(s) = lim s = .

s→0 s→0 τ s + 1 + KKth s 1 + KKth