Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

HTI., Elect. & Comp. Eng. Dept.

Control Systems_1; All groups

Profs. Waheed and Khaled Engs. Samir and Anas

Control Systems 1; Chapter 4

Problem 2: Derive the Laplace transform for the


following time functions: Find the output response, c(t),
for each of the systems shown in Figure P4.1. Also find
the time constant, rise time, and settling time for each
case. [Sections: 4.2, 4.3]
Answer
1 5 5 A B
a) C(s) =  c(t) = L−1(C(s))  Using partial fractions: = +
s s+5 s(s+5) s s+5

5 5 1 1
A=s | = 1 & B = (s + 5) | = −1  C(s) = −
s(s+5) s=0 s(s+5) s=−5 s s+5

1 1
 c(t) = L−1 ( ) − L−1 ( ) = 1 − e−5t; t ≥ 0 Or c(t) = [1 − e−5t]u(t)
s s+5

MATLAB code for partial fractions: num=5; den=[1 5 0]; [r , p , k]=residue(num , den)

Another method for determination of c(t):


𝐤
For the first-order system, whose closed loop transfer function is 𝐆(𝐬) = , the step
𝐬+𝐚
k k 5 5
k a a 5 5 1 1
response is the inverse Laplace transform of: C(s) = ≡ − = − = −
s(s+a) s s+a s s+5 s s+5
1 1
 c(t) = L−1 ( ) −1
−L ( )= 1−e −5t
; t ≥ 0 Or c(t) = [1 − e −5t ]
u(t)
s s+5

𝐤
For the first-order system, whose closed loop transfer function is 𝐆(𝐬) = and k=a= 5:
𝐬+𝐚

Time constant, Tc= 1/a= 1/5 seconds. Rise time, Tr= 2.2/a= 2.2/5= 0.44 seconds.

Settling time, Ts= 4/a= 4/5= 0.8 seconds.


1 20 20 A B
b) C(s) =  c(t) = L−1(C(s))  Using partial fractions: = +
s s+20 s(s+20) s s+20

20 20 1 1
A=s | = 1 & B = (s + 20) | = −1  C(s) = −
s(s+20) s=0 s(s+20) s=−20 s s+20

1 1
 c(t) = L−1 ( ) − L−1 ( ) = 1 − e−20t; t ≥ 0 Or c(t) = [1 − e−20t]u(t)
s s+20

MATLAB code for partial fractions: num=20; den=[1 20 0]; [r , p , k]=residue(num , den)
Eng. Samir A. Farag 1 Control Systems 1
Another method for determination of c(t):
𝐤
For the first-order system, whose closed loop transfer function is 𝐆(𝐬) = , the step
𝐬+𝐚
response is the inverse Laplace transform of:
k k 20 20
k a a 20 20 1 1
C(s) = ≡ − = − = −
s(s+a) s s+a s s+20 s s+20
1 1
 c(t) = L−1 ( ) − L−1 ( )= 1−e −20t
; t ≥ 0 Or c(t) = [1 − e −20t ]
u(t)
s s+20

𝐤
For the first-order system, whose closed loop transfer function is 𝐆(𝐬) = and k=a= 20:
𝐬+𝐚

Time constant, Tc= 1/a= 1/20 seconds. Rise time, Tr= 2.2/a= 2.2/20= 0.11 seconds.

Settling time, Ts= 4/a= 4/20= 0.2 seconds.


Problem 4: Find the capacitor voltage in the network shown in
Figure P4.2 if the switch closes at t = 0. Assume zero initial
conditions. Also find the time constant, rise time, and settling
time for the capacitor voltage. [Sections: 4.2, 4.3]
Answer
From the figure below
1
0.79s 1 1 5 0.7032
Vo (s) = Vi (s) × 1 = Vi (s) × = Vi (s) × = × . 1
= 0.7032
1.8+0.79s 1.8×0.79s+1 1.422s+1 s s+0.7032 1.422

k k
0.7032 k a a 1 1
Vo (s) = 5 ≡5 ≡ 5[ − ] = 5[ − ]
s(s+0.7032) s(s+a) s s+a s s+0.7032
1 1
 𝑣o (t) = 5L−1 ( ) − 5L−1 ( ) = 5 − 5e−0.7032t; t ≥ 0 Or c(t) = 5[1 − e−0.7032t]u(t)
s s+0.7032

Hint: you can use the partial fractions method introduced before to determine 𝒗𝐨 (𝐭), but for sure the
above method is too direct and simple, particularly for step response of first-order systems.

Time constant, Tc= 1/0.7032= 1.4221 seconds.

Rise time, Tr= 2.2/a= 2.2/0.7032= 3.1286 seconds.

Settling time, Ts= 4/a= 4/0.7032= 5.6883 seconds.


Problem 8: For each of the transfer functions shown below, find the locations of the poles
and zeros, plot them on the s-plane, and then write an expression for the general form of
the step response without solving for the inverse Laplace transform. State the nature of
each response (overdamped, underdamped, and so on).
[Sections: 4.3, 4.4]
2 5 10(s+7) 20
a) T(s) = b) T(s) = (s+3)(s+6) c) T(s) = (s+10)(s+20) d) T(s) = 2
s+2 s +6s+144
s+2 s+5
e) T(s) = f) T(s) = (s+10)2
s2 +9

Eng. Samir A. Farag 2 Control Systems 1


Answers
2 1 2 A B
a) T(s) =  step response C(s) = = +
s+2 s (s+2) s s+2

Poles: -2 Zeros: None

The general form of the step response without solving for the inverse Laplace transform:

c(t) = A + Be−2t

Nature of each response: first-order response.

MATLAB code to plot poles and zeros in s-domain

T=tf(2,[1 2]); % execute this command alone and look at the result

sgrid; pzmap(T)

5 1 5 A B C
b) T(s) = (s+3)(s+6)  step response C(s) = = + +
s (s+3)(s+6) s s+3 s+6

Poles: -3 and -6 Zeros: None

The general form of the step response without solving for the inverse Laplace transform:

c(t) = A + Be−3t + Be−6t

Nature of each response: overdamped response. See reference pages 169&172 for details

MATLAB code to plot poles and zeros in s-domain

T=tf(5 , conv([1 2] , [1 6])); sgrid; pzmap(T)

10(s+7) 1 10(s+7) A B C
c) T(s) = (s+10)(s+20)  step response C(s) = (s+10)(s+20)
= + +
s s s+10 s+20

Poles: -10 and -20 Zeros: -7

The general form of the step response without solving for the inverse Laplace transform:

c(t) = A + Be−10t + Be−20t

Nature of each response: overdamped response.

MATLAB code to plot poles and zeros in s-domain: T=tf([10 70] , conv([1 10] , [1 20])); sgrid; pzmap(T)

Eng. Samir A. Farag 3 Control Systems 1


20 1 20 A Bs+C
d) T(s) =  step response C(s) = = +
s2 +6s+144 s s2 +6s+144 s s2 +6s+144

Poles: −3 + j3√15 and −3 − j3√15 Hint: s 2 + 6s + 144 = (s + 3 + j3√15)(s + 3 − j3√15)

Zeros: None

The general form of the step response without solving for the inverse Laplace transform:

c(t) = A + Be(real part of complex pole pair)t cos((imaginary part of complex pole pair)t + some pahse)

= A + Be−3t cos(3√15t + ϕ).

Nature of each response: underdamped response.

MATLAB code to plot poles and zeros in s-domain: T=tf(20 , [1 6 144]); sgrid; pzmap(T)

s+2 1 s+2 A Bs+C


e) T(s) =  step response C(s) = = +
s2 +9 s s2 +9 s s2 +9

Poles: −j3 and j3 Hint: s 2 + 9 = (s + j3)(s − j3) Zeros: -2

The general form of the step response without solving for the inverse Laplace transform:

c(t) = A + Be(real part of complex pole pair)t cos((imaginary part of complex pole pair)t + some pahse)

= A + Be0 cos(3t + ϕ) = A + B cos(3t + ϕ).

Nature of each response: undamped response.

MATLAB code to plot poles and zeros in s-domain: T=tf(2 , [1 1 9]); sgrid; pzmap(T)

s+5 1 s+5 A B C
f) T(s) = (s+10)2  step response C(s) = (s+10)2
= + + (s+10)2
s s s+10

Poles: -10 and -10 Zeros: -5

The general form of the step response without solving for the inverse Laplace transform:
1
c(t) = A + Be−10t + Bte−10t Hint: L{te−at} = (s+a)2
Nature of each response: critically response. See solved exercises page 173 in reference

MATLAB code to plot poles and zeros in s-domain: T=tf([1 5] , conv([1 10] , [1 10])); sgrid; pzmap(T)

Eng. Samir A. Farag 4 Control Systems 1


Problem 20:
For each of the second-order systems that follow, find ζ, wn, Ts Tp, Tr, and %OS. [Section:
4.6].

16 0.04 1.05×107
a) T(s) = b) T(s) = c) T(s) =
s2 +3s+16 s2 +0.02s+0.04 s2 +1.6×103 s+1.05×107

Answers
16 w2n
a) T(s) = ≡
s2 +3s+16 s2 +2ζwn s+w2n

3
 wn2 = 16  wn = 4 rad/sec and 2ζwn = 2ζ(4) = 3  ζ = = 0.375
8

4 1 8 π π
Ts = = = sec. TP = = = 0.8472 sec.
ζwn ζ 3 wn √1−ζ2 4√1−0.3752

(0.375×π)
−ζπ/√1−ζ2 −
OS = e × 100 = e √1−0.3752 × 100 = 28.06 %
1
Tr = [1.76ζ3 − 0.417ζ2 + 1.039ζ + 1]
wn

1
= [1.76(0.375)3 − 0.417(0.375)2 + 1.039(0.375) + 1] = 0.356 sec.
4
0.04 w2n
b) T(s) = ≡
s2 +0.02s+0.04 s2 +2ζwn s+w2n

1
 wn2 = 0.04  wn = 0.2 rad/sec and 2ζwn = 2ζ(0.2) = 0.02  ζ = = 0.05
20

4 4 π π
Ts = = = 400 sec. TP = = = 15.73 sec.
ζwn 0.05×0.2 wn √1−ζ 2 0.2√1−0.052

(0.05×π)

−ζπ/√1−ζ2 √1−0.052
OS = e × 100 = e × 100 = 85.45 %
1
Tr = [1.76ζ3 − 0.417ζ2 + 1.039ζ + 1]
wn

1
= [1.76(0.05)3 − 0.417(0.05)2 + 1.039(0.05) + 1] = 5.26 sec.
0.2
1.05×107 w2n
c) T(s) = ≡  wn2 = 1.05 × 107
s2 +1.6×103 s+1.05×10 7 s2 +2𝛇wn s+w2n

20
 wn ≅ 3240 rad/sec and 2ζwn = 2ζ(3240) = 1.6 × 103  ζ = = 0.247
81

Eng. Samir A. Farag 5 Control Systems 1


4 4 π π
Ts = = ≅ 0.005 sec. TP = = ≅ 0.001sec.
ζwn 0.247×3240 wn √1−ζ 2 3240√1−0.2472
(0.247×π)
−ζπ/√1−ζ2 −
OS = e × 100 = e √1−0.2472 × 100 = 44.9 %
1
Tr = [1.76ζ3 − 0.417ζ2 + 1.039ζ + 1]
wn

1
= [1.76(0.247)3 − 0.417(0.247)2 + 1.039(0.247) + 1] = 3.88 × 10−4 sec.
3240
Problem 29: For each of the unit step responses shown in Figure P4.9, find the transfer
function of the system. [Sections: 4.3, 4.6].

Answers

From the response, we measure the time constant as depicted in Figure 4.5, that is, the time for
the amplitude to reach 63% of its final value. Since the final value is about 2, the time constant is
evaluated where the curve reaches 0.63 * 2= 1.26, or about 0.0244 second. Hence, a= 1/0.0244=
40.984. To find k, we realize from Eq. (4.11) that the forced response reaches a steady state
k
value of K/a=2. Substituting the value of a, we find = 2  k = 81.976
40.984
Eng. Samir A. Farag 6 Control Systems 1
cfinal ×a k 81.976
Thus, the transfer function for the system is G(s) = ≡ =
s+a s+a s+40.984
k k
k k
Hint: C(s) = ≡ a
− a
↔ c(t) = [1 − e−at]u(t) (4.11)
s(s+a) s s+a a
b) The response is for second-
order system whose transfer
function general form is:

k
G(s) =
s2 +2𝜁wn +w2n

w2n cfinal
= (1)
s2 +2𝜁wn +w2n

Evaluate the percent overshoot


and settling time, from which we
can find the poles and hence the
denominator.
cmax −cfinal
%OS =
cfinal

13.82−11.03
= ≅ 25.3% and from the figure, Ts = 2.62 sec.
11.03

Estimating a second-order system, we use Equation:


%OS 25.3
ln( 100 ) ln( 100 )
ζ=− %OS
=− 25.3
≅ 0.4 By MATLAB: -log(25.3/100)/sqrt(pi^2+ (log(25.3/100))^2)
√π2 +ln2 ( ) √π2 +ln2 ( )
100 100

4 4 4
Ts =  wn = = ≅ 3.82 rad/sec
ζ×wn ζ×Ts 0.4×2.62

The numerator can be found, as in the first-order system, from a knowledge of the measured and
expected steady-state values. From (1), wn2cfinal = 3.822 × 11.03 ≅ 160.95

Therefore, the system’s transfer function is:

w2n cfinal 160.95 160.95


G(s) = 2 ≡ =
s2 +2𝜁w n +wn s2 +2×0.4×3.82×𝑠+3.822 s2 +3.056𝑠+14.59

Another approximate method to find wn:


π π π
Tp = = 1  wn = = ≅ 3.43 rad/sec
wn √1−ζ2 √1−0.42 √1−0.4 2
Eng. Samir A. Farag 7 Control Systems 1
b) The response is for second-order system whose transfer function general form is:

k w2n cfinal
G(s) = = (1)
s2 +2𝜁wn +w2n s2 +2𝜁wn +w2n

From the graph, %OS = 40%. Using Eq. (4.39), ζ = 0.28. Also from the graph

cmax −cfinal 1.4−1


%OS = = = 40%
cfinal 1

%OS 40
ln( 100 ) ln(100)
ζ=− %OS
=− 40
= 0.28 By MATLAB: -log(40/100)/sqrt(pi^2+ (log(40/100))^2)
√π2 +ln2 ( ) √π2 +ln2 ( )
100 100

π
Tp = =4
wn √1−ζ2

π π
 wn = = = 0.8181 rad/sec By MATLAB: pi/(4*sqrt(1-0.28^2 ))
4√1−ζ2 4√1−0.28 2

From (1), wn2cfinal = 0.81812 × 1 ≅ 0.669

Therefore, the system’s transfer function is:

w2n cfinal 0.669 0.669


G (s ) = 2 ≡ =
s2 +2𝜁w n +wn s2 +2×0.28×0.8181×𝑠+0.81812 s2 +0.458𝑠+0.669

Another approximate method to find wn:


4 4 4
from the figure below, Ts = 130 sec. Ts =  wn = = ≅ 1 rad/sec
ζ×wn ζ×Ts 0.28×13

Where, settling time, Ts, is the time required for the transient’s damped oscillations to reach and
stay within ±2% of the steady-state value, i.e. 0.98 or 1.02 as shown in figure below:

Eng. Samir A. Farag 8 Control Systems 1


Monday, October 21, 2019

Samir A. Farag
 Teaching assistant at Electrical and computers Engineering department,
Higher Technological Institute (H.T.I), Industrial Area 2 – Next to Small
Industries Complex P.O. Box 228- 10th of Ramadan city, Egypt
 M.Sc. from Computer and Systems Engineering department, Faculty of
engineering, Zagazig University, Egypt, July 26, 2017 (Issue: Control of
Large Crane Structure)
 PhD program at Electrical Power and Machines Department, Faculty of
engineering, Cairo University.
 Facebook group: Mr. Eng. Samir A. Farag
 e-mails: samir.harb22@yahoo.com & samir.farag@hti.edu.eg
Eng. Samir A. Farag 9 Control Systems 1

You might also like