Control System Design

Assignment

Muhammad Kalim Ullah

May 7, 2017
Chapter 1

PID Controls and Two Degree

of Freedom

1.1 A type Problems

1.1.1 A-1
Describe briefly the dynamic characteristics of the PI controller, PD con-
troller, and PID controller.
Solution: The PI controller is characterized by the transfer function
1
Gc (s) = Kp (1 + ) (1.1)
Ti s
The PI controller is a lag compensator. It possesses a zero at s = T1 , and
a pole at s = 0. Thus,the characteristic of the PI controller is infinite gain
at zero frequency. This improves the steadystate characteristics. However,
inclusion of the PI control action in the system increases the type number
of the compensated system by 1, and this causes the compensated system
to be less stable or even makes the system unstable.Therefore, the values of
Kp and T , must be chosen carefully to ensure a proper transient response.
By properly designing the PI controller, it is possible to make the transient
response to a step input exhibit relatively small or no overshoot.The speed of
response, however, becomes much slower. This is because the PI controller,
being a low-pass filter, attenuates the high-frequency components of the sig-
nal.

1
The PD controller is a simplified version of the lead compensator.The PD
controller has the transfer function Gc (s), where
Gc (s) = Kp (1 + Td s) (1.2)
The value of Kp , is usually determined to satisfy the steady-state require-
ment.The corner frequency 1/Td , is chosen such that the phase lead occurs
in the neighborhood of the gain crossover frequency. Although the phase
margin can be increased, the magnitude of the compensator continues to
increase for the frequency region 1/Ki < . (Thus, the PD controller is a
high-pass filter.) Such a continued increase of the magnitude is undesirable,
since it amplifies high-frequency noises that may be present in the system.
Lead compensation can provide a sufficient phase lead, while the increase of
the magnitude for the high-frequency region is very much smaller than that
for PD control. Therefore, lead compensation is preferred over PD control.
Because the transfer function of the PD controller involves one zero, but no
pole, it is not possible to electrically realize it by passive RLC elements only.
Realization of the PD controller using op amps, resistors, and capacitors is
possible, but because the PD controller is a high-pass filter, as mentioned
earlier, the differentiation process involved may cause serious noise problems
in some cases.There is, however, no problem if the PD controller is realized
by use of the hydraulic or pneumatic elements.
The PD control, as in the case of the lead compensator, improves the transient-
response characteristics, improves system stability, and increases the system
bandwidth, which implies fast rise time.
The PID controller is a combination of the PI and PD controllers. It is a
lag-lead compensator. Note that the PI control action and PD control ac-
tion occur in different frequency regions. The PI control action occurs at
the low-frequency region and PD control action occurs at the high frequency
region.The PID control may be used when the system requires improvements
in both transient and steady-state performances.

1.1.2 A-2
Show that the transfer function U (s)/E(s) of the PID controller shown in
Figure 10-38 is
" #
U (s) T1 + T2 1 T1 T2 s
= K0 1+ + (1.3)
E(s) T1 (T1 + T2 )s (T1 + T2 )

2
 Figure 1.1: PID Controller

Assume that the gain K is very large compared with unity, or K  1.

Solution:
U (s) K
=  
E(s) 1 T1 s 1
1 + K K 1+T s 1+T s
0 1 2
K
=  
1 T1 s 1
K K 1+T s 1+T s
0 1 2
K (1 + T1 s)(1 + T2 s)
= 0 (1.4)
T1 s
  
1
= K0 1 + 1 + T2 s
T1 s
 
1 T2
= K0 1 + T2 s + +
T1 s T1
 
T1 + T2 1 T1 T2 s
= K0 1+ +
T1 (T1 + T2 )s T1 + T2

1.1.3 A-3
Consider the electronic circuit involving two operational amplifiers shown in
Figure 1.2. This is a modified PID controller in that the transfer function
involves an integrator and a first-order lag term. Obtain the transfer function
of this PID controller.

3
 Figure 1.2: Modified PID Controller

Solution: Since
1 R + R3 + R1 R3 C1 s
Z1 = + R3 = 1 (1.5)
1 +C s 1 + R1 C1 s
R1 1

and
1
Z2 = R2 + (1.6)
C2 s
we have
E(s) Z (1 + R2 C2 )(1 + R1 C1 s)
= 2 = (1.7)
Ei (s) Z1 C2 s(R1 + R3 + R1 R3 C1 s)
Also
Eo (s) R
= 5 (1.8)
E(s) R4
Consequently

Eo (s) Eo (s) E(s) R5 (R1 C1 s + 1)()R2 C2 s + 1)

= =  
Ei (s) E(s) Ei (s) R4 (R1 + R3 )C2 R1 R3
s R +R C1 s + 1
1 3
  
(1.9)
s + R 1C s + R 1C
R R
= 5 2 1 1 2 2

R4 R3 R +R
s s + R 1R C3
1 3 1

4
 Figure 1.3: Approximate Differentiator

Notice that R1 C1 and R2 C2 determine the locations of the zeros of the

controller, while R1 , R3 , and C1 affect the location of the pole on the negative
real axis. R5 /R1 adjusts the gain of the controller.

1.1.4 A-4
In practice, it is impossible to realize the true differentiator. Hence, we
always have to approximate the true differentiator Td s by something like

Td s
(1.10)
1 + Td s

One way to realize such an approximate differentiator is to utilize an inte-

grator in the feedback path. Show that the closed-loop transfer function of
the system shown in Figure 1.3 is given by the preceding expression. (In
the commercially available differentiator, the value of y may be set as 0.1).
Solution: The closed-loop transfer function of the system shown in Figure
1.3 is
1
C(s)
= (1.11)
R(s) 1 + T1 s
d
Note that such a differentiator with first-order delay reduces the bandwidth
of the closed-loop control system and reduces the detrimental effect of noise
signals.

5
1.1.5 A-5
Consider the system shown in Figure 1.4. This is a PID control of a second-
order plant G(s). Assume that disturbances D(s) enter the system as shown
in the diagram. It is assumed that the reference input R(s) is normally
held constant, and the response characteristics to disturbances are a very
important consideration in this system.

Figure 1.4: PID Controlled System

Design a control system such that the response to any step disturbance will
be damped out quickly (in 2 to 3 sec in terms of the 2% settling time).
Choose the configuration of the closed-loop poles such that there is a pair
of dominant closed-loop poles. Then obtain the response to the unit-step
disturbance input. Also, obtain the response to the unit-step reference input.
Solution: The PID controller has the transfer function
K(as + 1)(bs + 1)
Gc (s) = (1.12)
s
For the disturbance input in the absence of the reference input, the closed-
loop transfer function becomes
Cd (s) s
=
D(s) s(s2 + 3.6s + 9) + K(as + 1)(bs + 1) (1.13)
s
=
3 2
s + (3.6 + Kab)s + (9 + Ka + Kb)s + K
Hence we require
3.6 + Kab = 14
9 + Ka + Kb = 56 (1.14)
K = 160

6
which yields
ab = .065, a + b = 0.29375 (1.15)
The PID controller now becomes

K abs2 + (a + b)s + 1
 
Gc (s) =
s
2
160(0.065s + .29375s + 1
= (1.16)
s
2
10.4(s + 4.5192s + 15.385
=
s
With this PID controller the response to the disturbance is given by
s
Cd (s) = D(s)
3 2
s + 14s + 56s + 160
s (1.17)
= D(s)
2
(s + 10)(s + 4s + 16)

Clearly, for a unit-step disturbance input, the steady-state output is zero,

since

s2 1
lim cd (t) = lim sCd (s) = lim =0
t s s (s + 10)(s2 + 4s + 16) s

The response to a unit-step disturbance input can be obtained easily with

MATLAB. MATLAB Program produces a response curve as shown in Figure
1.5. From the response curve, we see that the settling time is approximately
2.7 sec.The response damps out quickly.Therefore, the system designed here
is acceptable. For the reference input r(t), the closed-loop transfer function
is
Cr (s) 10.4(s2 + 4.5192s + 15.385)
=
R(s) s3 + 14s2 + 56s + 160
(1.18)
10.4s2 + 47s + 160
=
s3 + 14s2 + 56s + 160
The response to a unit-step reference input can also be obtained by use of
MATLAB Program.The resulting response curve is shown in Figure 1.6.The
response curve shows that the maximum overshoot is 7.3% and the settling
time is 1.2 sec. The system has quite acceptable response characteristics.

7
MATLAB PROGRAM
numd = [0 0 1 0];
dend = [1 14 56 160];
t = 0:0.01:5;
[c1, x1, t] = step(numd, dend, t);
plot(t,c1)
grid
title(Response to Unit-Step Disturbance Input)
xlabel(t Sec)
ylabel(0utput to Disturbance Input)
figure
numr = [0 10.4 47 160];
denr = [1 14 56 160];
[c2, x2, t] = step(numr, denr, t);
plot(t,c2)
grid
title(Response to Unit-Step Reference Input)
xlabel(t Sec)
ylabel(0utput to Reference Input)

Figure 1.5: Response to Unit step disturbance input

8
 Figure 1.6: Response to unit step reference input

1.1.6 A-6
Consider the system shown in Figure 1.7. It is desired to design a PID
controller G
c (s) such that the dominant closed-loop poles are located at
s = 1 + j 3. For the PID controller, choose a = 1and then determine the
values of K and b. Sketch the root-locus diagram for the designed system.
Solution: Since
(s + 1)(s + b) 1
Gc (s)G(s) = K (1.19)
s s1 + 1

the sum of the angles at s = 1 + j 3, one of the desired closed-loop poles,
from the zero at s = 1 and poles at s = 0, s = j, and s = j is

90 143.794 120 110.104 = 283.898 (1.20)

Hence the zero at s = -b must contribute 103.898 .This requires that the
zero be located at
b = .5714 (1.21)

9
 Figure 1.7: PID-Controlled System

The gain constant K can be determined from the magnitude condition.

(s + 1)(s + .5714) 1 =1
|K | (1.22)
s s2 + 1 s=1+j 3
or
K = 2.333 (1.23)
Then the compensator can be written as follows:

(s + 1)(s + .5714
Gc (s) = 2.3333 (1.24)
s
The open-loop transfer function becomes

2.3333(s + 1)(s + .5714) 1

Gc (s)G(s) = (1.25)
s s2 + 1
From this equation a root-locus plot for the compensated system can be
drawn. Figure 1.8 is a root-locus plot.

C(s) 2.3333(s + 1)(s + .5714)

= (1.26)
R(s) s3 + s + 2.3333(s + 1)(s + .5714)

The closed-loop poles are located at s = 1 j 3 and s = -0.3333. A
unit-step response curve is shown in Figure 1.9. The closed-loop pole at s =
-0.3333 and a zero at s = -0.5714 produce a long tail of small amplitude.

10
 Figure 1.8: PID-Controlled System

Figure 1.9: Unit-step response of the compensated system

11
 Figure 1.10: I-PD Controlled System

Figure 1.11: PID Controlled System

1.1.7 A-7
Show that the I-PD-controlled system shown in Figure 1.10 is equivalent to
the PID-controlled system with input filter shown in Figure 1.11.
Solution : The closed-loop transfer function C(s)/R(s) of the I-PD
controlled system is
Kp
C(s) Ti s Gp (s)
=   (1.27)
R(s) 1
1 + Kp 1 + T s + Td s Gp (s)
i
The closed-loop transfer function C(s)/R(s) of the PID-controlled system

12
with input filter shown in Figure 1.11 is
 
1
Kp 1 + T s + Td s Gp (s)
C(s) 1 i
=  
R(s) 1 + Ti s + Ti Td s
1 + Kp 1 + T1s + Td s Gp (s)
i (1.28)
Kp
Ti s Gp (s)
= 
1
1 + T s + Td s Gp (s)
i
The closed-loop transfer functions of both systems are the same. Thus, the
two systems are equivalent.

1.1.8 A-8
The basic idea of the I-PD control is to avoid large control signals (which
will cause saturation phenomenon) within the system. By bringing the pro-
portional and derivative control actions to the feedback path, it is possible
to choose larger values for Kp and Td than those possible by the PID control
scheme.
Compare, qualitatively, the responses of the PID-controlled system and I-
PD-controlled system to the disturbance input and to the reference input.
Solution: Consider first the response of the I-PD-controlled system to the
disturbance input. Since, in the I-PD control of a plant, it is possible to
select larger values for Kp and Td than those of the PID-controlled case, the
I-PD-controlled system will attenuate the effect of disturbance faster than
the PID-controlled case.
Next, consider the response of the I-PD-controlled system to a reference in-
put. Since the I-PD-controlled system is equivalent to the PID-controlled
system with input filter (refer to Problem A-7), the PID-controlled system
will have faster responses than the corresponding I-PD-controlled system,
provided a saturation phenomenon does not occur in the PID-controlled sys-
tem.

1.1.9 A-9
In some cases it is desirable to provide an input filter as shown in Figure 1.12.
Notice that the input filter Gf (s) is outside the loop. Therefore, it does not

13
affect the stability of the closed loop portion of the system. An advantage of
having the input filter is that the zeros of the closed-loop transfer function
can be modified (canceled or replaced by other zeros) so that the closed loop
response is acceptable.
Show that the configuration in Figure 1.12 can be modified to that shown in

Figure 1.12: Block diagram of control system with input filter

Figure 1.13: modified block diagram

Figure 1.13, where Gd (s) = [Gf (s) l]Gc (s). The compensation structure
shown in Figure 1.13 is sometimes called command compensation.
Solution: For the system of Figure 1.12, we have

C(s) Gc (s)Gp (s)

= Gf (s) (1.29)
R(s) 1 + Gc (s)Gp (s)

14
for the system of Figure 1.13, we have
U (s) = Gd (s)R(s) + Gc (s)E(s)
E(s) = R(s) C(s)
C(s) = Gp (s)U (s)
Thus
(1.30)
C(s) = Gp (s)Gd (s)R(s) + Gc (s)[R(s) C(s)]
or
C(s) [G (s) + Gc (s)]Gp (s)
= d
R(s) 1 + Gc (s)Gp (s)
By substituting Gd (s) = [Gf (s) l]Gc (s) into Equation 1.30, we obtain

C(s) [Gf (s)Gc (s) Gc (s) + Gc (s)]Gp (s)

=
R(s) 1 + Gc (s)Gp (s)
(1.31)
Gc (s)Gp (s)
= Gf (s)
1 + Gc (s)Gp (s)
which is the same as Equation 1.29. Hence, we have shown that the systems
shown in Figures 1.12 and 1.13 are equivalent.
It is noted that the system shown in Figure 1.13 has a feedforward controller
Gd (s). In such a case, Gd (s) does not affect the stability of the closed-loop
portion of the system.

1.1.10 A-10
A closed-loop system has the characteristic that the closed-loop transfer func-
tion is nearly equal to the inverse of the feedback transfer function whenever
the open-loop gain is much greater than unity.
The open-loop characteristic may be modified by adding an internal feedback
loop with a characteristic equal to the inverse of the desired open-loop char-
acteristic. Suppose that a unity-feedback system has the open-loop transfer
function.
K
G(s) = (1.32)
(T1 s + 1)(T2 s + 1)

Determine the transfer function H(s) of the element in the internal feed-
back loop so that the inner loop becomes ineffective at both low and high

15
 Figure 1.14: control system

Figure 1.15: addition of the internal feedback loop to modify the closed loop
characteristic

frequencies.
Solution. Figure 1.14 shows the original system. Figure 1.15 shows the
addition of the internal feedback loop around G(s). Since
C(s) G(s) 1 G(s)H(s)
= = (1.33)
E(s) 1 + G(s)H(s) H(s) 1 + G(s)H(s)
if the gain around the inner loop is large compared with unity, then G(s)H(s)/[1+
G(s)H(s)] is approximately equal to unity, and the transfer function C(s)/E(s)
is approximately equal to l/H(s).
On the other hand, if the gain | G(s)H(s) | is much less than unity, the
inner loop becomes ineffective and C(s)/E(s) becomes approximately equal
to G(s).
To make the inner loop ineffective at both the low- and high-frequency ranges,
we require that
| G(j)H(j) | 1, for  1 and  1 (1.34)
Since, in this problem,
K
G(j) = (1.35)
(1 + jT1 )(1 + jT2 )

16
the requirement can be satisfied if H(s) is chosen to be
H(s) = ks (1.36)
because
Kkj
lim G(j)H(j) = lim =0
0 0 (1 + jT1 )(1 + jT2 )
(1.37)
Kkj
lim G(j)H(j) = lim =0
(1 + jT1 )(1 + jT2 )

Thus, with H(s) = ks (velocity feedback), the inner loop becomes ineffective
at both the low and high-frequency regions. It becomes effective only in the
intermediate-frequency region.

1.1.11 A-11
Consider the control system shown in Figure 1.16. This is the same system
as that considered in Example 10-1. In that example we designed a PID con-
troller Gc (s), starting with the second method of the Ziegler-Nichols tuning
rule. Here we design a PID controller using the computational approach with
MATLAB. We shall determine the values of K and a of the PID controller
(s + a)2
Gc (s) = K (1.38)
s
such that the unit-step response will exhibit the maximum overshoot between
10% and 2% (1.02 maximum output 1.10) and the settling time will be
less than 3 sec. The search region is
2 K , .05 a 2 (1.39)
Let us choose the step size for K to be 1 and that for a to be 0.05.
Write a MATLAB program to find the first set of variables K and n that will
satisfy the given specifications. Also, write a MATLAB program to find all
possible sets of variables K and a that will satisfy the given specifications.
Plot the unit-step response curves of the designed system with the chosen
sets of variables K and a.
Solution. The transfer function of the plant is
1
Gp (s) = (1.40)
s3 + 6s2 + 5s

17
The closed loop transfer function C(s)/R(s) is given by

C(s) Ks2 + 2Kas + Ka2

= (1.41)
R(s) s4 + 6s3 + (5 + K)s2 + 2Kas + Ka2
A possible MATLAB program that will produce the first set of variables I
and a that will satisfy the given specifications is given in MATLAB Program.
In this program we use two for loops. The specification for the settling time
is interpreted by the following four lines:

s = 501; while y(s) > .98 and y(s) < 1.02;

s = s 1; end;
(1.42)
ts = (s 1) .01;
ts < 3.0

Note that for t = 0 :0.01: 5, we have 501 computing time points. s = 501
corresponds to the last computing time point.
The solution obtained by this program is

K = 32, a = 0.2 (1.43)

with the maximum overshoot equal to 9.69% and the settling time equal to
2.64 sec.The resulting unit-step response curve is shown in Figure 1.17. Next,

Figure 1.16: control system

we shall consider the case where we want to find all sets of variables that
will satisfy the given specifications. A possible MATLAB program for this
purpose is given in MATLAB Program. Note that in the table shown in the
program, the last row should be ignored.(These are the last K and n values

18
for searching purposes.)
From the sort table, it seems that

K = 29, n = 0.25(maxovershoot = 9.52 (1.44)

and
K = 27, a = 0.2(maxovershoot = 5.5 (1.45)
are two of the best choices. The unit-step response curves for these two cases
are shown in Figure 1.18. From these curves, we might conclude that the
best choice depends on the system objective. If a small maximum overshoot
is desired, K = 27, a = 0.2 will be the best choice. If the shorter settling
time is more important than a small maximum overshoot, then K = 29, a =
0.25 will be the best choice.
MATLAB PROGRAM 1
t = 0:0.01:5;
for K = 50:-1:2;
for a = 2:-0.05:0.05;
num = [0 0 K 2 K a K a2 ];
den = [1 6 5 + K 2 K a K a2 ];
y = step(num, den, t);
m = max(y);
s = 501; while y(s) > 0.98 && y(s) < 1.02;
s = s 1; end;
ts = (s 1) 0.01;
if m < 1.10&&m > 1.02&&ts < 3.0
break;
end
end
if m < 1.10&&m > 1.02&&ts < 3.0
break
end
end
plot(t,y)
grid
title(Unit-Step Response)
xlabel(t sec)
ylabel(Output)
solution = [K; a; m; ts]

19
 Figure 1.17: Unit step response curve

solution =
32.0000
.2000
1.0969
2.6400
MATLAB PROGRAM 2
t = 0 : 0.01 : 5;
k = 0;
f ori = 1 : 49;
K(i) = 51 i 1;
f orj = 1 : 40;
a(j) = 2.05 j 0.05;
num = [0 0 K(i) 2 K(i) a(j) K(i) a(j) a(j)];
den = [1 6 5 + K(i) 2 K(i) a(j) K(i) a(j) a(j)];
y = step(num, den, t);
m = max(y);
s = 501; whiley(s) > 0.98&&y(s) < 1.02;
s = s 1; end;

20
 Figure 1.18: Unit step response curves

ts = (s 1) 0.01;
if m < 1.10&&m > 1.02&&ts < 3.0
k = k + 1;
table(k, :) = [K(i) a(j) m ts];
end
end
end
table(k, :) = [K(i) a(j) m ts];
sorttable = sortrows(table, 3);
K = sorttable(7, 1);
a = sorttable(7, 2);
num = [0 0 K 2 K a K a2 ];
den = [1 6 5 + K 2 K a K a2 ];
y = step(num, den, t);
plot(t, y)
grid
holdon
K = sorttable(2, 1);
num = [0 0 K 2 K a K a2 ];
den = [1 6 5 + K 2 K a K a2 ];
y = step(num, den, t);
plot(t, y)
title(0 U nit StepResponseCurves0 )

21
xlabel(0 t(sec)0 )
ylabel(0 0utput0 )
text(1.22, 1.22,0 K = 29, a = 0.250 )
text(1.22, 0.72,0 K = 27, a = 0.20 )

1.1.12 A-12
Consider the two-degrees-of-freedom control system shown in Figure 1.19.
The plant Gp (s) is given by

Figure 1.19: Two degree of freedom control system

100
Gp (s) = (1.46)
s(s + 1)

Assuming that the noise input N(s) is zero, design controllers Gc1 (s) and
Gc2 such that the designed system satisfies the following:

The response to the step disturbance input has a small amplitude and
settles to zero quickly(on the order of 1 sec to 2 sec).

The response to the unit-step reference input has a maximum overshoot

of 25% or less, and the settling time is 1 sec or less.

The steady-state errors in following ramp reference input and acceler-

ation reference input are zero.

22
Solution. The closed-loop transfer functions for the disturbance input and
reference input are given, respectively, by

Y (s) Gp (s)
=
D(s) 1 + Gc1 (s)Gp (s)
G (1.47)
Y (s) [Gc1 (s) + Gc2 (s)]Gp (s) c1
=
R(s) 1 + Gc1 (s)Gp (s)

Let us assume that Gcl (s) is a PID controller and has the following form:

K(s + a)2
Gc1 (s) = (1.48)
s
The characteristic equation for the system is

K(s + a)2 100

1 + Gc1 (s)Gp (s) = 1 + (1.49)
s s(s + 1)

Notice that the open-loop poles are located at s = 0 (a double pole) and s =
-1. The zeros are located at s = -a (a double zero).
In what follows, we shall use the root-locus approach to determine the values
of a and K. Let us choose the dominant closed-loop poles at s = -5 j5.Then,
the angle deficiency at the desired closed-loop pole at s = -5 + j5 is

135 135 128.66 + 180 = 218.66 (1.50)

The double zero at s = -a must contribute 218.66 . (Each zero must con-
tribute 109.33 .) By a simple calculation, we find

a = 3.2460 (1.51)

The controller Gc 1(s) is then determined as

K(s + 3.2460)2
Gc1 (s) = (1.52)
s
The constant K must be determined by use of the magnitude condition. This
condition is
| Gc1 (s)Gp (s) |s=5+j5 = 1 (1.53)

23
Since
K(s + 3.2460)2 100
Gc1 (s)Gp (s) = (1.54)
s s(s + 1)
we obtain
s2 (s + 1)
K =| |s=5+j5 = .11403 (1.55)
100(s + 3.2460)2
The controller Gc1 (s) thus becomes

.11403(s + 3.2460)2
Gc1 (s) =
s
2
.11403s + .7403s + 1.20148
= (1.56)
s
1.20148
= .7403 + + .11403s
s
Then, the closed-loop transfer function Y(s)/D(s) is obtained as follows:

Y (s) Gp (s)
=
D(s) 1 + Gc1 (s)Gp (s)
100
s(s+1)
= (1.57)
.11403(s+3.2460)2 100
1+ s s(s+1)
100s
=
s3 + 12.403s2 + 74.0284s + 120.148
The response curve when D(s) is a unit-step disturbance is shown in Figure
1.20.
Next, we consider the responses to reference inputs. The closed-loop transfer
function Y (s)/R(s) is

Y (s) [G (s) + Gc2 (s)]Gp (s)

= c1 (1.58)
R(s) 1 + Gc1 (s)Gp (s)

Let us define
Gc1 (s) + Gc2 (s) = Gc (s) (1.59)

24
 Figure 1.20: Step Response to disturbance input

Then
Y (s) Gc (s)Gp (s)
=
R(s) 1 + Gc1 (s)Gp (s)
(1.60)
100sGc (s)
=
s3 + 12.403s2 + +74.0284s + 120.148
To satisfy the requirements on the responses to the ramp reference input and
acceleration reference input, we use the zero-placement approach. That is,
we choose the numerator of Y(s)/R(s)to be the sum of the last three terms
of the denominator, or

100sGc (s) = 12.403s2 + 74.0284s + 120.148 (1.61)

from which we get

.12403s2 + .740284s + 1.20148

Gc (s) =
s (1.62)
1.20148
= .7403 + + .12403s
s

25
 Figure 1.21: Step Response to reference input

Hence, the closed-loop transfer function Y(s)/R(s) becomes as

Y (s) 12.403s2 + 74.0284s + 120.148

= (1.63)
R(s) s3 + 12.403s2 + 74.0284s + 120.148
The response curves to the unit-step reference input, unit-ramp reference in-
put, and unit acceleration reference input are shown in Figures 1.21,1.22 and
1.23 respectively.The maximum overshoot in the unit-step response is approx-
imately 25% and the settling time is approximately 1.2 sec. The steady-state
errors in the ramp response and acceleration response are zero. Therefore,
the designed controller Gc (s) given by Equation (1.62) is satisfactory.
Finally, we determine Gc2 (s). Noting that

Gc2 (s) = Gc (s) Gc1 (s) (1.64)

and from Equation 1.56

1.20148
Gc1 (s) = .7403 + + .11403s (1.65)
s

26
 Figure 1.22: ramp Response to reference input

Figure 1.23: acceleration Response to reference input

27
 Figure 1.24: Block Diagram of the designed system

we obtain
   
1.20148 1.20148
Gc2 (s) = .7403 + + .12403s .7403 + + .11403s
s s
= .01s
(1.66)
Equations 1.56 and 1.66 give the transfer functions of the controllers Gcl (s)
and Gc2 (s), respectively. The block diagram of the designed system is shown
in Figure 1.24.
Note that if the maximum overshoot were much higher than 25% and/or
the settling time were much larger than 1.2 sec, then we might assume a
search region (such as 3 a 6, 3 b 6, and 6 c 12) and use
the computational method presented in Example 10-4 to find a set or sets
of variables that would give the desired response to the unit-step reference
input.

1.2 B Type Problems

1.2.1 B-1
Consider the electronic PID controller shown in Figure 1.25. Determine the
values of R1 , R2 , R3 , R4 , C1 andC2 of the controller such that the transfer

28
 Figure 1.25: Electronic PID controller

function Gc (s) = Eo (s)/Ei (s) is

 
1
Gc (s) = 39.42 1 + + .7692s
3.077s
(1.67)
(s + .65)2
= 30.3215
s
Solution.
From this diagram we have,
R4 (R1 C1 + R2 C2 )
Kp = = 39.42
R3 R1 C2
Ti = R1 C1 + R2 C2 = 3.077 (1.68)
R1 C1 R2 C2
Td = = .7692
R1 C1 + R2 C2

So we notice that

R1 C1 + R2 C2 = 3.077(R1 C1 )(R2 C2 ) = .7692 3.077 = 2.3668 (1.69)

Hence we obtain,
R1 C1 = 1.5385, R2 C2 = 1.5385 (1.70)
Since we have six unknown variable and three equations, we can choose three
variables arbitrarily. So we choose C1 = C2 = 10F and one remaining

29
variable later. Then we get

R1 = R2 = 153.85k (1.71)

From the equation for Kp

f racR4 (R1 C1 + R2 C2 )R3 R1 C2 = 39.42

R4 (1.72)
= 3.42 .5 = 19.71
R3

We now choose R3 = 10k . Then R4 = 197.1

1.2.2 B-2
Consider the system shown in figure 1.26. Assume that disturbances D(s)
enter the system as shown in the diagram. Determine parameters K,a and b
such that the response to the unit-step disturbance input and the response
to the unit-step reference input satisfy the following specifications. The re-
sponse to the step disturbance input should attenuate rapidly with no steady
state error and the response to the step reference input exhibits a maximum
overshoot of 20% or less and a settling time of 2 sec. Solution.
For the reference input, the closed loop transfer function is

C(s) 2K(as + 1)(bs + 1)(s + 2)

= (1.73)
R(s) s(s + 1)(s + 10) + 2K(as + 1)(bs + 1)(s + 2)

Notice that the numerator is a polynomial in s of degree 3. In such a case, it

is advisable to reduce the degree of numerator polynomial by one by choosing

Figure 1.26: Control system

30
a=0. Then the closed loop transfer function becomes

C(s) 2k(bs + 1)(s + 2)

= (1.74)
R(s) s(s + 1)(s + 10) + 2K(bs + 1)(s + 2)

Let us choose the value of b to be .5 so that the zero of the controller is

located at s=-2. Then, the controller transfer functions Gc (s) becomes

K(.5s + 1) .5K(s + 2)
Gc (s) = = (1.75)
s s
Then
C(s) K(s + 2)2
= (1.76)
R(s) s(s + 1)(s + 10) + K(s + 2)2
The requirement on the response to the step disturbance input is that the
response should attenuate rapidly. Let us interpret this requirement to be
the settling time of 2 sec. By a simple MATLAB trail and error approach
on the value of K, we find that K=20 gives the settling time to be 2 sec. so
we choose K=20. with K=20, the closed loop transfer function CD (s)/D(s)
for the disturbance input becomes

CD (s) 2s2 + 4s
= (1.77)
D(s) s3 + 31s2 + 90s + 80
The following matlab program produces the response to the unit step distur-
bance input.
MATLAB PROGRAM 1
numd = [0240];
dend = [1319080];
step(numd, dend);
grid
title(0 U nitstepresponse(Disturbanceinput)0 );
This response curve corresponds to the settling time of 2 sec. This may not
be obvious so we plot unit ramp response
MATLAB PROGRAM 2
numdd = [00240];
dendd = [13190800];
step(numdd, dendd);
grid

31
Figure 1.27: Unit step response(Disturbance input)

Figure 1.28: Unit ramp response(Disturbance input)

32
 Figure 1.29: Unit step response(Reference input)

title(0 U nitrampresponse(Disturbanceinput)0 );
For the reference input the closed loop transfer function with K=20 is
C(s) 20s2 + 80s + 80
= (1.78)
R(s) s3 + 31s2 + 90s + 80
MATLAB PROGRAM 3
numr = [0208080];
denr = [1319080];
step(numr, denr);
grid
title(0 U nitstepresponse(Ref erenceinput)0 );
So the designed controller is
20(.5s + 1) 10(s + 2)
Gc (s) = = (1.79)
s s

1.2.3 B-3
Show that the PID controlled system shown in Figure 1.30 is equivalent
to I-PD controlled system with feedforward control shown in Figure 1.31.

33
Figure 1.30: PID Controller

Figure 1.31: I-PD controller

34
Solution. The closed loop transfer function of the system shown in Figure
1.30 is
Kp (1 + T1s + Td s)Gp (s)
C(s) i
= (1.80)
R(s) 1 + Kp (1 + T1s + Td s)Gp (s)
i
The closed loop transfer function of the system in Figure 1.31 can be obtained
as follows: Define the input to the block Gp (s) as U (s). Then,

Kp
U (s) = Kp (1 + Td s)R(s) + [R(s) C(s)] Kp (1 + Td s)C(s) (1.81)
Ti s
Also, we have
C(s) = Gp (s)U (s) (1.82)
Hence
C(s) 1 1
= Kp (1 + + Td s)R(s) Kp (1 + + Td s)C(s) (1.83)
Gp (s) Ti s Ti s

From which we obtain

Kp (1 + T1s + Td s)Gp (s)

C(s) i
= (1.84)
R(s) 1 + Kp (1 + T1s + Td s)Gp (s)
i

1.2.4 B-4
Consider the systems shown in Figure 1.32. Obtain the response of I-PD
controlled system to the unit step reference input with MATLAB. Solution.
We shall first obtain the closed loop transfer function of I-PD controlled
system.
C(s) 39.42
= (1.85)
R(s) s(s + 1)(s + 5) + 39.42(1 + .76925)
Where U(s) is the input to the minor loop. The open-loop transfer function
G(s) is
 
1 39.42
G(s) =
3.077s s(s + 1)(s + 5) + 39.42(1 + .76925)
(1.86)
12.8112
=
s4 + 6s3 + 35.3219s2 + 39.42s

35
The closed loop transfer function is

C(s) 12.8112
= (1.87)
R(s) s4 + 6s3 + 35.3219s2 + 39.42s + 12.8112

MATLAB PROGRAM num = [000012.8112];

den = [1635.321939.4212.8112];
t = 0 : .05 : 30;
step(num, den, t)
grid
title(0 U nitstepResponseof I P Dcontrolledsystem0 )

1.2.5 B-5
Check unit step disturbance input to systems in figure 1.32 and 1.34 and
show that they are same.
Solution. For PID controlled system in Figure 1.34, the closed loop transfer
function between the output and the disturbance input is

C(s) s
=
D(s) s2 (s + 1)(s + 5) + 39.42(s + .3250 + .7692s2 ) (1.88)
s
=
s4 + 6s3 + 35.3219s2 + 39.42s + 12.8112

Figure 1.32: I-PD Controller

36
Figure 1.33: Unit step response

Figure 1.34: PID Controller

37
 Figure 1.35: Step response to disturbance

For the I-PD controlled system in Figure 1.32, the closed loop transfer func-
tion between output and the disturbance input is

C(s) s
=
D(s) s2 (s + 1)(s + 5) + 39.42(s + .3250 + .7692s2 ) (1.89)
s
=
s4 + 6s3 + 35.3219s2 + 39.42s + 12.8112
Since the two closed loop transfer functions are identical, we get the same
unit step response curves for the two systems.
MATLAB PROGRAM
num = [00010];
den = [1635.321939.4212.8112];
step(num, den);
grid
title(0 Responsetounitstepdisturbanceinput0 )

The closed loop transfer function C(s)/R(s) of the system of Figure 1.34

38
is obtained as
1 + .7692s)
39.42(1 + 3.077s 1
C(s) s(s+1)(s+5)
=
R(s) 1 + .7692s)
1 + 39.42(1 + 3.077s 1
s(s+1)(s+5) (1.90)
30.3215s2 + 39.42s + 12.8112
=
s4 + 6s3 + 35.32186s2 + 3942s + 12.8112
The closed loop transfer function C(s)/R(s) of the system of figure 1.32 is
obtained as follows
39.42 1
C(s) s(s+1)(s+5)+39.42(1+.7692s) 3.077s
=
R(s) 1+ 39.42 1
s(s+1)(s+5)+39.42(1+.7692s) 3.077s (1.91)
12.8112
=
s4 + 6s3 + 35.32186s2 + 39.42s + 12.8112

1.2.6 B-6
Cosider the system shown in Figure 1.36. See if characteristic equation
changes with change in input.
Solution. For the reference input, the closed loop transfer function is

C(s) G1 (s)G2 (s)

= (1.92)
R(s) 1 + G1 (s)G2 (s)H(s)

For the disturbance input,

C(s) G2 (s)
= (1.93)
D(s) 1 + G1 (s)G2 (s)H(s

For the noise input,

C(s) G1 (s)G2 (s)H(s)

= (1.94)
N (s) 1 + G1 (s)G2 (s)H(s

Notice that characteristic equation of three closed loop system are the same

1 + G1 (s)G2 (s)H(s) = 0 (1.95)

39
Figure 1.36: Control system

Figure 1.37: Control system

40
1.2.7 B-7
Consider the system shown in figure 1.37. Find C(s)/R(s) and C(s)/D(s).
Solution. The closed loop transfer function C(s)/R(s) for the reference input

G1 G2 G3
C(s) 1+G2 G3 H2 G1 G2 G3
= = (1.96)
R(s) G1 G2 G3 1 + G2 G3 H2 + G1 G2 G3 H1
1 + 1+G
2 G3 H2
The closed loop transfer function C(s)/D(s) for the disturbance input is ob-
tain as follows: Noting that the feedforward transfer function is G3 (s) and
the feedback transfer function is [G1 (s)H1 (s) H2 (s)]G2 (s), and that the
closed loop system is a positive feedback system, we have

C(s) G3 G3
= = (1.97)
D(s) 1 G3 [G1 H1 H2 ]G2 1 + G1 G2 G3 H1 + G1 G3 H2

1.2.8 B-8
Consider the system shown in Figure 1.38 where K is an adjustable gain and
G(s) and H(s) are fixed components. The closed loop transfer function for
the disturbance is
C(s) 1
= (1.98)
D(s) 1 + KG(s)H(s)
to decrease disturbance K should be chosen as large as possible. is it the
same for Figure 1.39.
Solution. For the system in figure 1.39 the closed loop transfer function for
the disturbance input is

C(s) KG(s)H(s)
= (1.99)
D(s) 1 + KG(s)H(s)

To minimize the effect of disturbance K should be as small as possible. So,

answer is NO.

1.2.9 B-9
Show that the control systems shown in Figure 1.40 ,1.41,and 1.42 are two
degrees of freedom system in the diagram Gc1 (s) and Gc2 (s) are controllers

41
Figure 1.38: Control system with disturbance in feedforward path

Figure 1.39: Control system with disturbance in feedback path

42
Figure 1.40: Two degrees-of-freedom systems

Figure 1.41: Two degrees-of-freedom systems

Figure 1.42: Two degrees-of-freedom systems

43
and Gp (s) is the plant.
Solution.
Part a
Y (s) Gc1 Gp
=
R(s) 1 + Gc1 Gc2 Gp
Y (s) Gp
=
D(s) 1 + Gc1 Gc2 Gp
Y (s) Gc1 Gc2 Gp (1.100)
=
N (s) 1 + Gc1 Gc2 Gp
GY R = Gc1 GY D
G Gp
GY N = Y D
Gp
Part b
Y (s) G G Gp
= c1 c2
R(s) 1 + Gc2 Gp
Y (s) Gp
=
D(s) 1 + Gc2 Gp
Y (s) Gc2 Gp (1.101)
=
N (s) 1 + Gc2 Gp
GY R = Gc1 Gc2 GY D
G Gp
GY N = Y D
Gp
Part c
Y (s) Gc1 Gp
=
R(s) 1 + Gc2 Gp
Y (s) Gp
=
D(s) 1 + Gc2 Gp
Y (s) Gc2 Gp (1.102)
=
N (s) 1 + Gc2 Gp
GY R = Gc1 GY D
G Gp
GY N = Y D
Gp
In the three systems, if GY D is given, then GY N is fixed but GY R is not
fixed. Thus the systems are two-degree of freedom systems.

44
 Figure 1.43: Three degrees-of-freedom system

1.2.10 B-10
Show that the control system shown in Figure 1.43 is a three degree of free-
dom system. Solution. Define the input signal to box Gc3 as A(s). Then
for D(s)=0 and N(s)=0

A(s) = Gc2 R(s) + Gc1 [R(s) Y (s)] G1 Gc3 A(s)

Y (s) = Gc3 G1 G2 A(s)
(G G )R(s) Gc1 Y (s)
Y (s) = Gc3 G1 G2 c1 c2 (1.103)
1 + Gc3 G1
Y (s) (Gc1 + Gc2 )Gc3 G1 G2
=
R(s) 1 + Gc3 G1 + Gc1 Gc3 G1 G2

To Find Y/D, R(s)=N(s)=0

A(s) = Gc1 [Y (s)] G1 [D(s) + Gc3 A(s)]

Y (s) = G1 G2 [D(s) + Gc3 A(s)]
(1.104)
Y (s) G1 G2
=
D(s) 1 + Gc3 G1 + Gc1 Gc2 G1 G2

Next we shall find Y(s)/N(s). For R(s)=D(s)=0

A(s) = Gc1 [Y (s) + N (s)] G1 Gc3 A(s)

Y (s) = G1 G2 Gc3 A(s)
(1.105)
Y (s) Gc1 Gc3 G1 G2
=
N (s) 1 + Gc3 G1 + Gc1 Gc3 G1 G2

45
 Figure 1.44: Control system

so from above three equations we get

GY N = Gc1 Gc3 GY D
(1.106)
GY R = GY N + Gc2 Gc3 GY D

So We have three degree of freedom.

1.2.11 B-11
Consider the PID control in Figure 1.44. Unit step response should exhibit
10% overshoot and settling time of 2 sec. write a matlab program 1 K
4, .4 4. Plot unit step response curves. Solution. The open loop transfer
function of the system is

K(s + a)2 1.2

G(s) =
s (.3s + 1)(s + 1)(1.2s + 1)
1.2Ks2 + 2.4Kas + 1.2Ka2
=
.36s4 + 1.86s3 + 2.5s2 + s
C(s) 1.2Ks2 + 2.4Kas + 1.2Ka2
=
R(s) .36s4 + 1.86s3 + (2.5 + 1.2K)s2 + (1 + 2.4Ka)s + 1.2Ka2
(1.107)
The requirement in this problem is that the maximum overshoot in the unit
step response is that
p < .1, Mp > .02 (1.108)

46
 Figure 1.45: Unit step response

where Mp is the maximum overshoot. In terms of output y to the unit step

input,
m < 1.1, m > 1.02 (1.109)
where
m = max(y) (1.110)
MATLAB PROGRAM
t = 0 : 0.01 : 8;
f orK = 4 : .05 : 1;
f ora = 4 : .05 : .4;
num = [0 0 1.2 K 2.4 K a 1.2 K a2 ];
den = [.36 1.86 2.5 + 1.2 K 1 + 2.4 K a 1.2 K a2 ];
y = step(num, den, t);
m = max(y);
if m < 1.1&&m > 1.02
break;
end
end
if m < 1.1&&m > 1.02
break
end

47
end
plot(t, y)
grid
title(0 U nitstepresponse0 )
xlabel(0 tsec0 )
ylabel(0 Output0 )
sol = [K a m]
So we get K=4, a=.7 and m=1.0846

1.2.12 B-12
For same control system as in B-11. find K and a such that overshoot is
between 3 and 8 and settling time is less than 2%. choose 2 K 4, .5
a 3.
Solution.

K(s + a)2 1.2

G(s) =
s (.3s + 1)(s + 1)(1.2s + 1)
1.2Ks + 2.4Kas + 1.2Ka2
2
=
.36s4 + 1.86s3 + 2.5s2 + s
C(s) 1.2Ks2 + 2.4Kas + 1.2Ka2
=
R(s) .36s4 + 1.86s3 + (2.5 + 1.2K)s2 + (1 + 2.4Ka)s + 1.2Ka2
(1.111)
MATLAB PROGRAM
t=0:0.01:8;
for K=4:-.05:2;
for a=3:-.05:.5;
num = [0 0 1.2 K 2.4 K a 1.2 K a2 ];
den = [.36 1.86 2.5 + 1.2 K 1 + 2.4 K a 1.2 K a2 ];
y=step(num,den,t);
m=max(y);
s=801; while y(s).98 && y(s)1.02
s=s-1;end;
ts=(s-1)*.01;
if m1.08 && m1.03 && ts2
break;
end

48
 Figure 1.46: Unit step response

end
if m1.1 && m1.02 && ts2
break
end
end
plot(t,y)
grid
title(Unit step response)
xlabel(t sec)
ylabel(Output)
sol= [K a m ts]

1.2.13 B-13
Consider the two-degrees-of-freedom control system shown in Figure 1.47.
Design controllers such that unit step response to disturbance input goes to
zero quickly. Maximum overshoot is 25% and settling time is 2 sec Solution.

3(s + 5)
Gp (s) = (1.112)
s(s + 1)(s2 + 4s + 13)

49
 Figure 1.47: two degree of freedom control system

Closed loop transfer functions are

Y (s) Gp (s)
=
D(s) 1 + Gp (s)Gc1 (s)
(1.113)
Y (s) [G (s) + Gc2 (s)]Gp (s)
= c1
R(s) 1 + Gp (s)Gc1 (s)
Assume that Gc1 (s) is a PID controller with a filter and has the following
form:
K(s + a)2 s2 + 4s + 13
Gc1 (s) = (1.114)
s (s + 5)(s + 27)
The characteristic equation is

K(s + a)2 s2 + 4s + 13 3(s + 5)

1 + Gc1 (s)Gp = 1 +
s (s + 5)(s + 27) s(s + 1)(s2 + 4s + 13)
3K(s + a)2
=1+
s2 (s + 1)(s + 27)
(1.115)
we find that K=58 and a=1.4

58(s + 1.4)2 s2 + 4s + 13
Gc1 (s) =
s (s + 5)(s + 27)
(1.116)
58s2 + 162.4s + 113.68 s2 + 4s + 13
=
s (s + 5)(s + 27)

50
The closed loop transfer function is obtained as

3(s+5)
Y (s) s(s+1)(s2 +4s+13)
=
D(s) 358(s+1.4)2
1+ 2 (1.117)
s (s+1)(s+27)
3s3 + 96s2 + 405s
=
(s2 + 4s + 13)(s4 + 28s3 + 201s2 + 487.2s + 341.04)

Now finding Y/R

Y (s) (Gc1 + Gc2 )Gp Y (s)

= = (Gc1 + Gc2 ) (1.118)
R(s) 1 + Gc1 Gp D(s)

where
Gc = Gc1 + Gc2 (1.119)
So, we have
3 2 +405s
Gc 3s +96s
Y (s)
= s2 +4s+13 (1.120)
R(s) s4 + 28s3 + 201s2 + 487.2s + 341.04
To get zero error for ramp and acceleration input we have to put Numerator
of above equation equal to last three terms of denominator

3s3 + 96s2 + 405s

Gc = 201s2 + 487.2s + 341.04
2
s + 4s + 13
(201s2 + 487.2s + 341.04)(s2 + 4s + 13)
Gc (s) = (1.121)
3s3 + 96s2 + 405s
67s2 + 162.4s + 113.68 s2 + 4s + 13
=
s (s + 5)(s + 27)

Closed loop Y/R becomes

Y (s) 201s2 + 487.2s + 341.04

= (1.122)
R(s) s4 + 28s3 + 201s2 + 487.2s + 341.04

51
Now to find Gc2 (s)

Gc2 (s) = Gc (s) Gc1 (s)

67s2 + 162.4s + 113.68 s2 + 4s + 13
Gc (s) =
s (s + 5)(s + 27)
58s + 162.4s + 113.68 s2 + 4s + 13
2 (1.123)
Gc1 (s) =
s (s + 5)(s + 27)
2
s + 4s + 13
Gc2 (s) = 9s
(s + 5)(s + 27)

52