Automatic Control

Introduction

By Vasfi Emre Ömürlü, Ph.D.

Concepts of Automatic Control

Feedback Control Steady-state tracking

Open loop control Time domain analysis
Closed loop control Time domain specifications
Disturbances Poles and zeros
Differential equations (auxiliary - PID control – proportional,
reviewed) integral, derivative control
Laplace transform (auxiliary - reviewed) PID control parameters
Transfer functions Tuning methods of PID control
Block diagram representation of dynamic parameters
systems Ziegler-Nichols tuning method
Mathematical modeling of physical Root Locus
systems (auxiliary - reviewed) Gain
Mechanical, electrical, hydraulic systems
(auxiliary – reviewed)
Stability – BIBO stability, Routh-Hurwitz
criterion

2
By Vasfi Emre Ömürlü, Ph.D., 2020
Figure 1-1 (p. 2)
Basic components of a control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-2 (p. 3)
Idle-speed control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-3 (p. 4)
Industrial sewing machine.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-4 (p. 4)
Ideal velocity profile of one start-stop cycle of an industrial sewing machine.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-5 (p. 5)
Solar collector field.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-6 (p. 5)
Conceptual method of efficient water extraction using solar power.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-7 (p. 6)
Important components of the sun-tracking control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-8 (p. 6)
Elements of an open-loop control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-8 (p. 7)
Block diagram of a closed-loop idle-speed control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-10 (p. 7)
(a) Typical response of the open-loop idle-speed control system.
(b) Typical response of the closed-loop idle-speed control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-14 (p. 13)
Schematic diagram of a typical dc closed-loop system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-15 (p. 13)
Schematic diagram of a typical ac closed-loop control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-16 (p. 14)
Block diagram of a sampled-data control system.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Figure 1-17 (p. 14)
Digital autopilot system for a guided missile.

Automated Control Systems, 8/E by Benjamin C. Kuo and Farid Golnaraghi

Copyright © 2003 John Wiley & Sons. Inc. All rights reserved.
Automatic Control I
Mathematical Modeling of
Dynamic Systems
By Vasfi Emre Ömürlü, Ph.D.
Mechanical System Modeling
Electrical System Modeling
System Examples

18

By Vasfi Emre Ömürlü, Ph.D., 2005

 19

By Vasfi Emre Ömürlü, Ph.D., 2005

Introduction

To understand and control complex systems, one must

obtain quantitative mathematical models of systems.
Because the systems under consideration are dynamic
in nature, the descriptive equations are usually
differential equations.
Dynamic system problems
 Define the system and its components
 Formulate the mathematical model and list the necessary
assumptions
 Write the differential equations describing the model
 Solve the equations for the desired output variables
 Examine the solutions and the assumptions
 If necessary, reanalyze or redesign the system
20
By Vasfi Emre Ömürlü, Ph.D., 2005
Some dynamic systems and their
mathematical representations

b f(t) k b  y  k  ( y  x)  f (t )
m
m  x  k  ( x  y )  0
y x

m  x  b  x  k  x  f (t )

Valve to adjust the

flow rate Q3 between
Automatic control valve to adjust the tanks
liquid levels of the tanks by
controlling the flap angle 

Q1 q1
dh1
A1  Q1  a1 h1  a3 h1  h2
dt
Discharge Discharge
h1 valve dh
Q2
valve h2 A2 2   a2 h2  a3 h1  h2
Q3 Q4 dt
21

By Vasfi Emre Ömürlü, Ph.D., 2005

Disk Drive Read System

22
By Vasfi Emre Ömürlü, Ph.D., 2005
Mechanical Accelerometer

23
By Vasfi Emre Ömürlü, Ph.D., 2005
Mechanical System Elements

Inductive storage

Capacitive storage

Energy dissipator

24
By Vasfi Emre Ömürlü, Ph.D., 2005
Force-spring system.

25

By Vasfi Emre Ömürlü, Ph.D., 2005

Torque torsional spring system.

26

By Vasfi Emre Ömürlü, Ph.D., 2005

Dashpot for viscous friction.

27

By Vasfi Emre Ömürlü, Ph.D., 2005

Frictional relationships of linear and
nonlinear frictional forces.
(a) Viscous friction. (b) Static friction. (c)
Coulomb friction.

28

By Vasfi Emre Ömürlü, Ph.D., 2005

Force-mass system.

29

By Vasfi Emre Ömürlü, Ph.D., 2005

Torque-inertia system.

30

By Vasfi Emre Ömürlü, Ph.D., 2005

Rotary-to-linear motion control system
(lead screw).

31

By Vasfi Emre Ömürlü, Ph.D., 2005

Rotary-to-linear motion control system
(rack and pinion).

32

By Vasfi Emre Ömürlü, Ph.D., 2005

Rotary-to-linear motion-control system
(belt and pulley).

33

By Vasfi Emre Ömürlü, Ph.D., 2005

Gear train.

34

By Vasfi Emre Ömürlü, Ph.D., 2005

Physical model of backlash between two
mechanical elements.

35

By Vasfi Emre Ömürlü, Ph.D., 2005

Input-output characteristic of backlash.

36

By Vasfi Emre Ömürlü, Ph.D., 2005

Mechanical System Modeling

37
By Vasfi Emre Ömürlü, Ph.D., 2005
(a) Mass-spring-friction system. (b)
Free-body diagram (c) State diagram.

38

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Mass-spring-friction system.
(b) Free-body diagram. (c) State
diagram.

39

By Vasfi Emre Ömürlü, Ph.D., 2005

Electric network analogous to the
mechanical system in previous page.

40

By Vasfi Emre Ömürlü, Ph.D., 2005

Example

Find the transfer function

between force input and the
output of velocities v1 and v2.
show that electrical analogous is
represented by the same
differential equation.

41
By Vasfi Emre Ömürlü, Ph.D., 2005
Rotational system

42

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Motor-load system. (b) Free-body
diagram.

43

By Vasfi Emre Ömürlü, Ph.D., 2005

State diagram of the system in previous
page

44

By Vasfi Emre Ömürlü, Ph.D., 2005

Gear train with friction and inertia.

45

By Vasfi Emre Ömürlü, Ph.D., 2005

 46

By Vasfi Emre Ömürlü, Ph.D., 2005

 47

By Vasfi Emre Ömürlü, Ph.D., 2005

 48

By Vasfi Emre Ömürlü, Ph.D., 2005

 49

By Vasfi Emre Ömürlü, Ph.D., 2005

 50

By Vasfi Emre Ömürlü, Ph.D., 2005

 51

By Vasfi Emre Ömürlü, Ph.D., 2005

 52

By Vasfi Emre Ömürlü, Ph.D., 2005

 53

By Vasfi Emre Ömürlü, Ph.D., 2005

 54

By Vasfi Emre Ömürlü, Ph.D., 2005

 55

By Vasfi Emre Ömürlü, Ph.D., 2005

 56

By Vasfi Emre Ömürlü, Ph.D., 2005

Electrical System Elements

Inductive storage

Capacitive storage

Energy dissipator

57
By Vasfi Emre Ömürlü, Ph.D., 2005
Electrical System Modeling

Find the transfer function between v1 and v2.

58
By Vasfi Emre Ömürlü, Ph.D., 2005
Electrical System Examples

Using Kirchoff’s Voltage Law for both loops

I1
( s )  ?  vin  v R1  v R2  v L
Vin
di1
vin  R1  i1  R2  (i1  i 2 )  L For left loop
dt
1
c
0 i 2 .dt  R2 .(i 2  i1 ) For right loop

Applying  Laplace  Transform

Above electrical circuit has two loops in which  1 
R2  I 1 ( s )    R2   I 2 ( s )
the left loop has the system input Vin and the  Cs 
right loop the output current i2. The transfer
function between input and output is required. I ( s )   1  R2 Cs   I ( s )
Two voltage equations using Kirchoff’s
1  R Cs  2
 2 
Voltage Law will be written.
v ≡ voltage difference between two ends of the element (V), i ≡ current throught the element (A),
R ≡ resistance (Ω,ohm), L ≡ inductance (H, henry), C ≡ capacitance (F, farad)
59

By Vasfi Emre Ömürlü, Ph.D., 2005

Electrical System Modeling

60
By Vasfi Emre Ömürlü, Ph.D., 2005
(a) RLC network.
(b) State diagram.

61

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Network
(b) State diagram.

62

By Vasfi Emre Ömürlü, Ph.D., 2005

Electrical System Examples

Vin ( s )  R1  I 1 ( s )  R2  I 1 ( s )  I 2 ( s )  Ls  I 1 ( s )
Vin ( s )  ( R1  R2 )  Ls   I 1 ( s )  R2  I 2 ( s )
 1  R2 
Vin ( s )  ( R1  R2 )  Ls      I 2 ( s )  R2  I 2 ( s )
 2R Cs 
  1  R2 Cs  
Vin ( s )  ( R1  R2 )  Ls      R2   I 2 ( s )
  R2 Cs  
LR2 Cs 2  R1  R2 R2 Cs  Ls  R1  R2   CR2 s
2

Vin ( s )   I 2 (s)
CR2 s
LR2 Cs 2  R1 R2 C  L s  R1  R2 
Vin ( s )   I 2 ( s)
CR2 s
I 2 (s) CR2 s

Vin ( s ) LR2 Cs 2  R1 R2 C  L s  R1  R2 
System transfer function

63

By Vasfi Emre Ömürlü, Ph.D., 2005

Ten-turn rotary potentiometer (courtesy
of Helipot Division of Beckman
Instruments, Inc.

64

By Vasfi Emre Ömürlü, Ph.D., 2005

Linear motion potentiometer with built-in
operational amplifier (courtesy of Waters
Manufacturing, Inc.)

65

By Vasfi Emre Ömürlü, Ph.D., 2005

Electric circuit representation of a
potentiometer.

66

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Potentiometer used as a position
indicator.
(b) Two potentiometers used to sense
the positions of two shafts.

67

By Vasfi Emre Ömürlü, Ph.D., 2005

Block diagram representation of
potentiometer arrangements in Fig. 4-
25.

68

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) A dc-motor,
position-control
system with
potentiometers as
error sensors. (b)
Typical waveforms of
signals in the control
system of part (a).

69

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) An ac-control
system with
potentiometers as
error detectors.
(b) Typical
waveforms of
signals in the
control system of
part (a).

70

By Vasfi Emre Ömürlü, Ph.D., 2005

Velocity-control system with tachometer
feedback.

71

By Vasfi Emre Ömürlü, Ph.D., 2005

Position-control system with tachometer
feedback.

72

By Vasfi Emre Ömürlü, Ph.D., 2005

Rotary incremental encoder (courtesy of
DISC Instruments, Inc.).

73

By Vasfi Emre Ömürlü, Ph.D., 2005

Linear incremental encoder (courtesy of
DISC Instruments, Inc.).

74

By Vasfi Emre Ömürlü, Ph.D., 2005

Typical incremental optomechanics.

75

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Typical rectangular output waveform of a
single-channel encoder device (bidirectional).
(b) Typical dual-channel encoder signals in
quadrature (bidirectional).

76

By Vasfi Emre Ömürlü, Ph.D., 2005

(a) Typical sinusoidal output waveform of a
single-channel encoder device.
(b) Typical dual-channel encoder signals in
quadrature.

77

By Vasfi Emre Ömürlü, Ph.D., 2005

One cycle of the output signals of a
dual-channel incremental encoder.

78

By Vasfi Emre Ömürlü, Ph.D., 2005

PM DC Motor Modeling

Normally, field (stator) and armature (rotor)

windings are present in electrical motors.
Controlling through-current and excitation
voltage on both windings one can change the
electrical field on both and control either
rotaal velocity or torque generated.
Depending on the connection type of these
field and armature windings, DC motors vary
like:
Shunt (parallel) connected
Two pole simple representation of a Permanent Series connected
Magnet DC Motor. Stator pole are permanent Differentially connected
magnets meaning that constant magnetic field Compound connected
is generated. Varying pole locations Etc.
electrically on rotor windings by changing the
electrical current direction, rotor will rotate to
balance itself with the opposite stator pole.

79

By Vasfi Emre Ömürlü, Ph.D., 2005

Torque
production in a
dc motor.

80

By Vasfi Emre Ömürlü, Ph.D., 2005

Cross-section
view of a
permanent-
magnet (PM)
iron-core dc
motor.

81

By Vasfi Emre Ömürlü, Ph.D., 2005

Cross-section
view of a
surface-wound
permanent-
magnet (PM)
dc motor.

82

By Vasfi Emre Ömürlü, Ph.D., 2005

Cross-section
view of a
surface-wound
permanent-
magnet (PM)
dc motor.

83

By Vasfi Emre Ömürlü, Ph.D., 2005

Cross-section side view of a moving-coil
dc motor.

84

By Vasfi Emre Ömürlü, Ph.D., 2005

Cross-section
view of a
brushless,
permanent-
magnet (PM)
iron-core dc
motor.

85

By Vasfi Emre Ömürlü, Ph.D., 2005

Model of a separately excited dc motor.

86

By Vasfi Emre Ömürlü, Ph.D., 2005

State diagram of a dc-motor system.

87

By Vasfi Emre Ömürlü, Ph.D., 2005

Block diagram of a dc-motor system.

88

By Vasfi Emre Ömürlü, Ph.D., 2005

PM DC Motor Modeling

Vex: excitation voltage to excite the

system [V]
ia: armature current which flows
through armature windings [A]
Ra: armature resistance [ohm]
La: armature inductance [H]
vbemf: back electromotive force (voltage
load on the armature windings because of the
rotational speed of the motor) [V]
Kv : velocity constant of the motor
which can be seen on any DC motor catalog.
Constant which relates rotational velocity to
back emf voltage [V/(rad/sec)]
Kt: torque constant of the motor
Model contains electrical and mechanical parts which relates the armature current to the
since the motor have both electrical and generated torque
mechanical components. System is excited [Nm/A]
from electrical part with a current or voltage J: motor inertia [kgm2]
input and then, this electrical excitation is B: bearing damping [Nm/(rad/sec)]
converted to mechanical rotational energy. It TL: load torque [Nm]
rotates the mechanical inertial element with 89
some bearings (damping) on it. By Vasfi Emre Ömürlü, Ph.D., 2005
 PM DC Motor Modeling

di a
v ex  Ra  i a  La   v bemf Differential equation of the electrical part dynamics
dt 
K v ( t )

J 
  K t  i a  B    TL Differential equation of the mechanical part dynamics
Vex ( s )  Ra  I a ( s )  La  s  I a ( s )  K v  ( s ) Taking the Laplace transform of above equations
J  s  ( s )  K t  I a ( s )  B  ( s )  T L ( s ) assuming zero initial conditions
Vex ( s )  ( Ra  La  s )  I a ( s )  K v  ( s )
(J  s  B) 1
I a ( s)   ( s )  TL ( s )
Kt Kt
 (J  s  B) 1 
Vex ( s )  ( Ra  La  s )    ( s )  T L ( s )   K v  ( s )
 Kt Kt 
 (J  s  B)  ( Ra  La  s )
Vex ( s )  ( Ra  La  s )   K v   ( s )  TL ( s )
 K t  K t

 JL s 2  BLa  JRa s  BRa  K v K t   ( Ra  La s )

Vex ( s )   a   ( s )  TL ( s )
 K t  K t
( s ) ( s )
Vex ( s ) TL ( s )
   
Kt ( Ra  La s )
( s )  Vex ( s )  TL ( s )
JLa s  BLa  JRa s  BRa  K v K t 
2
JLa s  BLa  JRa s  BRa  K v K t 
2
  90  
transfer  function between excitation voltage transfer  function between load  torque
and  motor  velocity and  motor  velocity
By Vasfi Emre Ömürlü, Ph.D., 2005
PM DC Motor Open Loop Block
Diagram

Above equations of PM DC motor can be

represented with this block diagram. When
you back-solve this block diagram you will get
the same transfer function found previously

91

By Vasfi Emre Ömürlü, Ph.D., 2005

PM DC Motor Control Problem

Why do we want to apply control for a DC motor?

1. Speed control
2. Improve system performance
3. Reduce the effect of disturbance – load torque

92

By Vasfi Emre Ömürlü, Ph.D., 2005

Maxon PM DC motor
http://www.maxonmotorusa.com/

93

By Vasfi Emre Ömürlü, Ph.D., 2005

Maxon PM DC motor
http://www.maxonmotorusa.com/

94

By Vasfi Emre Ömürlü, Ph.D., 2005

Maxon PM DC motor
http://www.maxonmotorusa.com/

95

By Vasfi Emre Ömürlü, Ph.D., 2005

Various Systems

96
By Vasfi Emre Ömürlü, Ph.D., 2005
Various Systems

97
By Vasfi Emre Ömürlü, Ph.D., 2005
Electric Motor Traction Control

98
By Vasfi Emre Ömürlü, Ph.D., 2005
Nonlinear control system.

99

By Vasfi Emre Ömürlü, Ph.D., 2005

Magnetic-ball-suspension system.

100

By Vasfi Emre Ömürlü, Ph.D., 2005

Systems with transportation lag.

101

By Vasfi Emre Ömürlü, Ph.D., 2005

Schematic diagram of a sun-seeker
system.

102

By Vasfi Emre Ömürlü, Ph.D., 2005

Coordinate system of the sun-seeker
system.

103

By Vasfi Emre Ömürlü, Ph.D., 2005

Nonlinear characteristic of the error
discriminator. The abscissa is tan α, but
it is approximated by α for small values
of α.

104

By Vasfi Emre Ömürlü, Ph.D., 2005

Block diagram of the sun-seeker
system.

105

By Vasfi Emre Ömürlü, Ph.D., 2005

 106

By Vasfi Emre Ömürlü, Ph.D., 2005

 107

By Vasfi Emre Ömürlü, Ph.D., 2005

 108

By Vasfi Emre Ömürlü, Ph.D., 2005

 109

By Vasfi Emre Ömürlü, Ph.D., 2005

 110

By Vasfi Emre Ömürlü, Ph.D., 2005

 111

By Vasfi Emre Ömürlü, Ph.D., 2005

 112

By Vasfi Emre Ömürlü, Ph.D., 2005

 113

By Vasfi Emre Ömürlü, Ph.D., 2005

 114

By Vasfi Emre Ömürlü, Ph.D., 2005

 115

By Vasfi Emre Ömürlü, Ph.D., 2005

 116

By Vasfi Emre Ömürlü, Ph.D., 2005

 117

By Vasfi Emre Ömürlü, Ph.D., 2005

 118

By Vasfi Emre Ömürlü, Ph.D., 2005

 Laplace Transform:
Properties
Theorems
Partial Fraction Expansion

119

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform

It is advantageous to solve linear diff. equations by…

By using, we can convert many common functions into
algebraic equations of a complex variable “s”.
Operations like differentiation and integration can be replaced
by algebraic equations.
A linear differential equations can be transformed into an
algebraic equation.
If the algebraic equation in s is solved for the dependent
variable, then the solution of the differential equation may be
found by use of a Laplace Transform Table or by use of the
Partial Fraction Expansion Technique.

120
By Vasfi Emre Ömürlü, Ph.D., 2005
Advantage of the Laplace Transform

It allows the use of graphical techniques for predicting the

system performance without actually solving system diff.
equations.
When we solve the differential equation, both the transient
and steady-state component of the solution can be obtained
simultaneously.

121
By Vasfi Emre Ömürlü, Ph.D., 2005
Some dynamic systems and their
mathematical representations

b f(t) k b  y  k  ( y  x)  f (t )
m
m  x  k  ( x  y )  0
y x

m  x  b  x  k  x  f (t )

Valve to adjust the

flow rate Q3 between
Automatic control valve to adjust the tanks
liquid levels of the tanks by
controlling the flap angle 

Q1 q1
dh1
A1  Q1  a1 h1  a3 h1  h2
dt
Discharge Discharge
h1 valve dh
Q2
valve h2 A2 2   a2 h2  a3 h1  h2
Q3 Q4 dt
122

By Vasfi Emre Ömürlü, Ph.D., 2005

Complex Variable

A complex number has real and imaginary parts s    j

which are both constants. When both of these are
variables, it is said to be complex variable.
G ( s )  Gx  jG y
A complex function G(s) has a real part and
G ( s )  Gx  G y
2 2
imaginary part where Gx and Gy are real quantities.
Gy
G ( s )  tan 1
Gx
Complex functions are commonly encountered in
linear control systems and are single valued
functions of s.

123
By Vasfi Emre Ömürlü, Ph.D., 2005
Euler`s Theorem

2 4 6
cos   1     ...
2! 4! 6!
3 5 7
sin        ...
3! 5! 7!

( j ) 2 ( j ) 3 ( j ) 4
cos   j sin   1  ( j )     ...
2! 3! 4!

( x) 2 ( x)3 ( x) 4
Since e  1  ( x) 
x
   ...
2! 3! 4!

cos   j sin   e j Euler’s theorem

Also cos   j sin   e  j

124
By Vasfi Emre Ömürlü, Ph.D., 2005
(Inverse) Laplace Transform
definitions and its existence

f(t)=a function of time t such that f(t)=0 for t<0

s=a complex variable

L=Laplece transform notation
F(s)=Laplace transform of f(t) L f(t)  F ( s)   f (t )  e  st dt
0
The reverse process of Laplace Transform is also possible and is called Inverse Laplace Transform
with notation L-1
Generally, we do not use this integral function to find Laplace Transform of a function. We use
simpler methods.

The Laplace Transform of a function f(t) exists if the Laplace integral converges which will converge
if f(t) is sectionally continuous in every finite interval in the range t>0 and if it is of exponential
order as t approaches infinity.

125

By Vasfi Emre Ömürlü, Ph.D., 2005

Some Common Laplace Transforms
as examples…
Step function
0 for t  0
f(t)  
step
A for t  0
7
signal strength (signal unit)

6

L f (t )   A  e dt 
5
4
A  st
3
2
1 0
s
0
0 2 4 6 8 10 12
time(sec)

The step function whose height is unity is called unit-step function. The unit-step function that
occurs at t=to is frequently written as 1(t-to).

126
By Vasfi Emre Ömürlü, Ph.D., 2005
Some Common Laplace Transforms
as examples…
Exponential function
 0 for t  0
f(t)    at
 for t  0
exp decay

 A e 6

signal strength (signal unit)

5
 
L f (t )   A  e
4
 at
 e dt  A   e
 st ( s  a ) t
dt 3

1
0 0
0

L f (t ) 
A 0 2 4 6
time(sec)
8 10 12

sa

127
By Vasfi Emre Ömürlü, Ph.D., 2005
Some Common Laplace Transforms
as examples…
Ramp function

 0 for t  0
f(t)  
At for t  0
 
L f (t )  A  t  e  st dt  A  te  st dt
 
ramp

signal strength (signal unit)

0 0 6

 
5
 st  st
e e 4

 At 
s 
 A
 s
dt 3
2
0 0 1

 0
A

 st 0 2 4 6 8 10 12
 A
s
 e dt  2
time(sec)

0
s

128
By Vasfi Emre Ömürlü, Ph.D., 2005
Some Common Laplace Transforms
as examples…
Sinusoidal function

 0 for t  0
f(t)   recall A sin t  
A j t
e  e  j t 
 A  sin t for t  0 2j
A  ( j  s ) t 
 
L f (t ) 
A
2 j 0
e 
j t
 e  j t
e  st
dt 
2 j  0
  e dt  0

e  ( j  s ) t
dt 

A  1 1  A
   
2 j  s  j s  j  s 2   2
sine

signal strength (signal unit)

4

-2

-4

-6
0 2 4 6 8 10 12
time(sec)

129
By Vasfi Emre Ömürlü, Ph.D., 2005
Most Used Laplace Transformations

f(t) F(s)
A A/ s
At A / s2
At n A  n! / s n 1
A  e  at A /( s  a )
A  sin( a  t ) A  [a /( s 2  a 2 )]
A  cos( a  t ) A  [ s /( s 2  a 2 )]
A  e bt sin( a  t ) A  a /[s  b   a 2 ]
2

A  e bt cos( a  t ) A  ( s  b) /[s  b   a 2 ]

2

130

By Vasfi Emre Ömürlü, Ph.D., 2005

Shapes of mostly used signals

8 At n
A
signal strength (signal unit)

6 At
A  e  at
4

-2
 bt A  e bt sin( a  t )
Ae cos( a  t )
-4
A  cos( a  t ) A  sin( a  t )
-6
0 2 4 6 8 10 12
time(sec)

131

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform properties –
superposition

f(t)    f1 (t )    f 2 (t )

L  f1 (t )    f 2 (t )     f1 (t )    f 2 (t ) e  st  dt
0
 
    f1 (t )  e  st  dt     f 2 (t )  e  st  dt

0
  
0
 
L  f1 ( t ) L  f 2 ( t )

   F1 ( s )    F2 ( s )

Scaling property: L  f1 (t )    F1 ( s)

132

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform properties – time
delay

133

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform properties – time
delay

Suppose f(t) is delayed by λ>0. The Laplase transform of the function,


L f1 (t )   f (t   )  e  st  dt
0
Define a new variable, t’=t- λ, and then, dt’=dt, f(t)=0 for t<0
  f(t)

 f (t ' )  e  dt '   f (t ' )  e

 s (t '  )  st '  s
F1 ( s )   e  dt '
 0 const A/t0

 e  s   f (t ' )  e  st '  dt '  e  s  F ( s )

0

L  f ( t ) F ( s )

L f (t   )  e  s  F ( s)
0 t0 λ λ+t0

134

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform properties –
differentiation

d 
L  f (t )  sF ( s )  f (0)
 dt 
 d2 
L  2 f (t )  s 2 F ( s )  sf (0)  f (0)
 dt 
 dn  n
L  n f (t )  s F ( s )  s n 1 f (0)  s n  2 f (0)  ........  sf ( n  2 ) (0)  f ( n 1) (0)
 dt 

f(0) is the initial condition of the function and df(0)/dt is the initial condition of the derivative of
the function. For instance, if f(t) is representing the position of a mechanical system, position
initial condition will be f(0) and velocity initial condition will be df(0)/dt.

135

By Vasfi Emre Ömürlü, Ph.D., 2005

Some Common Laplace Transforms
as examples…

Pulse function
f(t)

A
 for 0  t  t0  A A
f(t)   t0  f (t )  1(t )  1(t  t0 )
t  t0 
A/t0  0 t0 t0
for t  0,
A  
 
A   st 
L f (t )    1(t )  1(t  t0 )   e dt     1(t )  e  st dt   1(t  t0 )  e  st dt 
A
0 0 
t t0 t0  0 0

A A  st0 A
  e  (1  e  st0 )
t0 s t0 s t0 s
t
0 t0

Here, A and t0 are all constants. The pulse function may be considered as a step function of
height A/t0 that begins at t=0 and that is superimposed by a negative step function of height
A/t0 beginning at t=t0.

136

By Vasfi Emre Ömürlü, Ph.D., 2005

Some Common Laplace Transforms
as examples…

Impulse function  A
 lim for 0  t  t0
f(t)
f(t)  t0 0 t0
 0 for t  0, t  t0

L f (t )  lim
A
A/t0 (1  e  st0 )
t0 0 t s
0

d
dt

A(1  e  st0 ) 
 lim 0  As / s  A
d
t0 s 
t 0 0

t dt0
0,t0

Since the height of the impulse function is A/t0 and duration is t0 , the area under the impulse is
A. when t0 approaches 0, area remains A. Note that the magnitude of the impulse is measured
by its area.
The impulse function whose area is equal to unity is called unit-impulse function or the Dirac
delta function.
137

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform theorems – Final
Value Theorem
The final value theorem relates the steady-state behavior of f(t) to the behavior of sF(s) in the
neighborhood of s=0. This theorem applies if and only if lim f(t) exists when t approaches
infinity which means that f(t) settles down to a definite value for t = infinity. If all poles of sF(s)
lie in the left half s plane (complex plane), this limit exists.

lim f (t )  lim sF ( s )
t  s 0

By using this method, we obtain the scalar value of a already-Laplace-transformed function at

t=infinity without transforming back to the time domain.
Example: find the steady-state value of the following function.
3( s  2)
Y (s)  poles  are  on  LHP
s ( s 2  2s  10)
 
 2  4  410
poles:
2

3( s  2)
lim y (t )  lim sY ( s )  lim s  3/ 5
t  s 0 s 0 s( s  2s  10)
2
138

By Vasfi Emre Ömürlü, Ph.D., 2005

Laplace Transform theorems – Initial
Value Theorem and DC Gain
By using this theorem, we are able to find the value of f(t) at t=0+ directly from the Laplace
transform of f(t). The initial value theorem is stated as: if f(t) and its derivative are both
transformable

f (0)  lim sF ( s)
s 
  
should  exist

DC gain is the ratio of the output of a system to its input after all transient have decayed. G(s) is
the transfer function of a system which defines the input-output dynamics of that system in s-
domain.

DC  Gain  lim G ( s )
s0

139

By Vasfi Emre Ömürlü, Ph.D., 2005

Partial Fraction Expansion

Why do we need?
Inverse Laplace Transform can be applied by using inversion integral theorem, yet
this method would be cumbersome. Thus, it will be easier, generally, to split a function in s-
domain to simpler forms whose transformations can be reached from Laplace tables.

called
th m 
zeros
mdeg ree . polynomial. with .m. zi .roots

B( s ) b1  s m  b2  s m 1  ......  bm 1  ( s  zi )
F ( s)   n 1
 K i n1
s  a1  s  ......  an
n


A( s )
 (s  p j )
n th deg ree. polynomial. with .n. p j .roots j 1

called
poles

 F1 ( s )  F2 ( s )  F3 ( s )  ......  L1F ( s )  L1F1 ( s ) L1F2 ( s ) L1F3 ( s ) ......

f (t )  f1 (t )  f 2 (t )  f 3 (t )  .......
Depending on the roots of the denominator of the function in s-domain which is called “poles”
there are three cases for partial fraction expansion:
1. Denominator has distinct roots, 2. Denominator has complex roots, 3. Denominator has
repeated roots.
140

By Vasfi Emre Ömürlü, Ph.D., 2005

Partial Fraction Expansion – Distinct
Poles
If poles are distinct, above equation can be separated into n parts…
C1 C2 Cn
F ( s)    ...  , for C1
( s  p1 ) ( s  p 2 ) (s  pn )
 C1 C2 Cn 
( s  p1 )  F ( s)  ( s  p1 )     ...  
 ( s  p1 ) ( s  p 2 ) (s  pn ) 
 
 ( s  p1 )  C 2 ( s  p1 )  C n 
( s  p1 )  F ( s) s  p  C1   ...    C1
1
 ( s  p ) ( s  p ) 
 2
 n
 0 0  s  p
1

C1  ( s  p1 )  F ( s ) s  p
1

General solution if all poles are distinct.

If some of the poles are distinct, same formula
is used to find the “C” constants. C n  (s  p n )  F (s) s  p
n

141

By Vasfi Emre Ömürlü, Ph.D., 2005

Partial Fraction Expansion –Distinct
Complex Poles
If some or all poles are complex…
1 C C s  C3 1
F (s)   1  22 , C1  s  F ( s ) s 0  1
s ( s  s  1)
2
s
 s  s 1 ( s  s  1) s 0
2 1,2

solve.as
usual

1 C2 s  C3 (C2  1) s 2  (C3  1) s  1
1
1
   
s ( s 2  s  1) s s 2  s  1 s s2  s 1  
0,8
arrange
  
s 1

signal strength (signal unit)

1
(C2  1) s 2  (C3  1) s  1  1  C2  1, C3  1  F ( s )   2
s s  s 1 0,6 step
sine decay

s 1/ 2 1/ 2 s 1/ 2 1 3/2 cosine decay

 2
 2
 2 0,4
f(t)

 3  3 3  3
( s  1 / 2) 2   
 ( s  1 / 2) 2    ( s  1 / 2)  
2

 2   2   2 
 0,2

1 s 1/ 2 1 3/2
F (s)   2
 2
s  3 3  3 0

( s  1 / 2)  
2

 ( s  1 / 2)  
2


 2   2 
-0,2
t 3 1 t 2 3 0 2 4 6 8 10
f (t )  1  e 2 cos t e sin t time(sec)
2 3 2
142

By Vasfi Emre Ömürlü, Ph.D., 2005

Partial Fraction Expansion –
Repeated Poles
1,2
If some or all poles are repeated…

s 2  2s  3 C1 C2 C3
F (s)    
1

( s  1) 3
s  1 s  1 s  13
2

s 2  2s  3 3 C C3 
s  1 3
 s  1  1 
C2
 3
 s  1 C1  s  1C2  C3
2 0,8

signal strength (signal unit)

 s  1 s  1 s  1 
f(t)
( s  1) 3 2
e^-t

s  13 F ( s) s1  C3  2, also d

ds
  
s  13 F ( s)  d s  12 C1  s  1C2  C3
ds
 0,6
t^2*e^-t

d
 

 s  1 F ( s )   C2  2s  2 s1  0
 ds
3

 s1
0,4

differentiating again
1  d2
2  ds
 3


 2 s  1 F ( s )   C1 
1
2 1 0,2

 s1
2
1 0 2
F (s)     f (t )  e t  t 2 e t
s  1 s  1 s  1
0
2 3
0 2 4 6 8 10
time(sec)

143

By Vasfi Emre Ömürlü, Ph.D., 2005

Example: Tank Dynamics

A process tank dynamics is given as ;

H ( s) 1

Q( s ) s  10
Q(s) varies in repeatedly as;

-Find h(t) in response to q(t)

-Discuss about the magnitude of h(t) at t = 105

144

By Vasfi Emre Ömürlü, Ph.D., 2005

Example: Tank Dynamics

1
H (s)   Q( s ) We need to find Q(s) in order to solve for H(s) then for h(t)
s  10
For the first ramp in I
I
In order to turn the
first ramp in I into a
step. Will be effective
after T seconds II
In order to turn the
step in II into a
negative ramp in III.
Will be effective after
2T seconds
III
In order to turn the
negative ramp in III IV
into 0 in IV. Will be
effective after 3T
seconds
145

By Vasfi Emre Ömürlü, Ph.D., 2005

Example: Tank Dynamics

Q( s) 
Q0 / T
s2
Q0 / T
 e  sT 
s2
Q /T
 e  2 sT  02
s
Q /T
 e 3 sT  02
s
1/s2 is the common term for every
sub-input. Other terms are
constants and time delay factors.
So, solve only for 1/s2 and then
146
apply for all system.
By Vasfi Emre Ömürlü, Ph.D., 2005
Example: Tank Dynamics

1 C1 C 2 C3
H ( s)    
s  10  s 2 s s 2 s  10
1 1
C 3  ( s  10) H ( s )  
s  10 s2 s  10 100
1 1
C 2  s 2 H ( s)  
s 0 s  10 s 0 10
d  d  1  1 1
C1   s 2 H ( s)     
 ds  s 0 ds  s  10  s 0 s  10
2
s 0 100
 1 / 100 1 / 10 1 / 100
H (s)   2 
s s s  10

147

By Vasfi Emre Ömürlü, Ph.D., 2005

Example: Tank Dynamics

1 t 1 10t 1 1
h(t )    e  h(t )   t  e 10t
100 10 100 10 10
h(t) for only 1/s^2

0,9

0,8

0,7

0,6

0,5

This result is only for 1/s2 0,4

input, yet other input 0,3

responses can be 0,2

generated by this response 0,1
because of the 0
superposition and scaling 0 0,2 0,4 0,6 0,8 1
properties of Laplace time (sec)

Transform
148

By Vasfi Emre Ömürlü, Ph.D., 2005

 Example: Tank Dynamics

Overall system response is

   Q0 / T Q / T 10t 
   (Q0 / T )t  0 e  for t T
  10 10 
   Q0 / T  (Q / T )t  Q0 / T e 10t  
  10 0
10 
   " T  t  2T
   Q0 / T  (Q / T )(t  T )  Q0 / T e 10( t T ) 
  10 0
10 
  Q /T Q / T 10t 
 0
 (Q0 / T )t  0 e  
  10 10

   Q0 / T Q0 / T 10(t T )  tank height, h(t)
  (Q0 / T )(t  T )  e  " 2T  t  3T
  10 10 
h(t )   0,35

   Q / T Q / T 

0
 (Q0 / T )(t  2T )  0 e 10( t  2T )  0,3
  10 10  0,25
  Q / T Q / T 10t 

height (m)
 0
 (Q0 / T )t  0 e  
0,2

 10 10
 0,15
  Q0 / T
 
Q0 / T 10(t T )
 (Q0 / T )(t  T ) 
0,1
 e

 10 10
 " 3T  t  
0,05

  Q0 / T Q0 / T 10(t  2T )  0
  (Q0 / T )(t  2T )  e  0 0,5 1 1,5
10 10 
 time (sec)

  Q0 / T Q0 / T 10(t 3T ) 
 (Q / T )(t  3T )  e 
 10 0
10  149

By Vasfi Emre Ömürlü, Ph.D., 2005

Example: A Mass-Damper-Spring
System

?
System mathematical model

x(0)  0,05m mx  bx  kx  f (t )

x (0)  0
f (t )  1N
m  1kg
b  30 N / m / sec 1  sT 1  sT 1 3 sT 1  4 sT 1 5 sT 1
F (s)   e  e  e  e  e
k  200 N / m s2 s2 s2 s2 s2 s2
x(t )  ?
150

By Vasfi Emre Ömürlü, Ph.D., 2005

Example – Dynamic System
Response by Laplace Transform

t
G(s)=1/(s+2) ?
Dynamic response of the
1 0  t 1 system above is required
u (t)=  against a second longing Y (s)  G ( s)  U ( s)
0 1 t  
pulse function.

 1   1 1 s 
    e 
1  1  s s  2 s s 
u ( s)   e
s  s 

151

By Vasfi Emre Ömürlü, Ph.D., 2005

 Example – Dynamic System
Response by Laplace Transform

 1   1   c1   c 2 
Y (s)          
s  2 s  s  2  s  1

c1  ( s  2)  Y ( s) s 2  0,5
t

c2  s  Y ( s) s 0  0,5 Response of the System to a

second long pulse

 0,5   0,5 
Y (s)  
0,5
 

system output

0,4
s  2  s  0,3

 1   1  5
0,2
Y (s)     e One-second 0,1
 s  2  s delayed of YI(s). 0
0 1 2 3

 0,5e  0,5

2 t
0  t 1 time (sec)
y (t )  
 0,5e  2t  0,5  0,5e  2(t 1)  0,5 1  t   152

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-1 (Time delay)

 1 , 0  t 1
 
u(t)   0 , 1  t  3 
 1 , 3  t   
 

L u(t)  ?

153

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-1
L f (t   )  e  s  F ( s)

 1 
 s 0 
 
 1 
U(s)    e  s  1 
 s 
 1  s 
 s e 3 
 

1 1  1 
U(s)      e  s     e 3s
s s  s 154

By Vasfi Emre Ömürlü, Ph.D., 2005

Ex-2 (Differentation)

 
y  10 y  9y  5t
y(0)  1

y(0)  2

Find the Laplace Transform of this

equation…

155

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-2

d 
L  f (t )  sF ( s )  f (0)
 dt 
 d2 
L  2 f (t )  s 2 F ( s )  sf (0)  f (0)
 dt 
 dn 
L  n f (t )  s n F ( s )  s n 1 f (0)  s n  2 f (0)  ........  sf ( n  2 ) (0)  f ( n 1) (0)
 dt 

156

By Vasfi Emre Ömürlü, Ph.D., 2005

 System Response
12

10

Ex-2


6

magnitude
 
L y  s  Y(s)  s  y(0)  y(0)
2 4

L y  s  Y(s)  y(0)

2

0

L y  Y(s)
-2
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
time (sec)

5
L 5t 
s2 5
Y(s) s  10s  9   s  12  2
2

s
5 s  12
Y(s)  2  2
s (s  9)(s  1) s (s  9)(s  1)
157

By Vasfi Emre Ömürlü, Ph.D., 2005

Ex-3 (Distinct Poles)

s2
F(s) 
(s  9)(s  1)

Find the Inverse Laplace Transform of this

equation…

158

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-3

s2 A B
F(s)   
(s  9)(s  1) s  9 s  1

C n  (s  p n )  F (s) s  p
n

11
A  (s  9)F(s) s  9 
8
3
B  (s  1)F(s) s 1  159
8
By Vasfi Emre Ömürlü, Ph.D., 2005
 Impulse Response
20

18

16

14

Ex-3 12

Amplitude
10

3
6

11 4

F(s)  8  8 2

s  9 s 1
0 0.05 0.1 0.15 0.2 0.25 0.3
Time (sec)

f (t)  L F(s)
1

11 9 t 3 t
f (t)  e  e
8 8
160

By Vasfi Emre Ömürlü, Ph.D., 2005

Ex-4 (Repeated poles)

s  25
F(s)  2 )
s (s  5)

Find the Inverse Laplace Transform of this

equation…

s  25 A B C
F(s)  2 )  2 
s (s  5) s s s5
161

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-4
30 6
C   (s  5)F(s) s 5  
25 5
B  (s2 )F(s)  s  0  5
d 2  d  s  25 
A   (s )F(s)    
 ds s  0 ds  s  5 s  0
d  s  25   (s  5)  (s  25)  30 6
     
ds  s  5 s  0  (s  5) 2
s  0 25 5

162

By Vasfi Emre Ömürlü, Ph.D., 2005

 Impulse Response
12

10

Ex-4

Amplitude
6

6 6
5 5 2

F(s)   2 5
s5
0

s s 0 0.05 0.1 0.15 0.2 0.25

Time (sec)
0.3 0.35 0.4 0.45 0.5

f (t)  L F(s)
1

6 6 5t
f (t)   5t  e
5 5

163

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5

L1/L2 =1/2
M=1 kg
R=60Nsec/m
K=800N/m
F(t)=1 N
- Find y(t)

Y(s) L1 1
  2
F(s) L2 ms  rs  k

164

By Vasfi Emre Ömürlü, Ph.D., 2005

Ex-5

 F0   F0   F0   F0 
 T  T  T  T
F(s)     e  st    e 2st    e 3st  
s2 s2 s2 s2
165

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5
Y(s) L1 1
  2
F(s) L2 ms  rs  k
L1 1
Y(s)   2  F(s)
L2 ms  rs  k
1 1 1 Common term for every sub-input
Y(s)   2  2
2 s  60s  800 s

1 1 1 C1 C2 C3 C4
Y(s)   2  2   2  
2 s  60s  800 s s s (s  40) (s  60)
166

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5
1
C2  s2 Y(s)  s  0 
1600
d  d 1 
C1    s2  Y(s)     2 
 ds   ds s  60s  800 s  0
 2s  60 
 2   9  10 5

 (s  60s  800) s  0
2

 1 
C3  (s  40)Y(s) s 40   2   1,56 105
 2s (s  60) s 40
 1 
C4  (s  60)Y(s) s 60   2   6, 9  10 6

 2s (s  40)  s 40
167

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5

9  105 6, 25  104 1,56  105 6, 9 106

Y(s)    
s s 2
s  40 s  60
5 t 6 t
y(t)  9  105  6, 25  104  t  e 1,5610  e 6,910

168

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5

Overall system response:

1,56105 t 6,9106 t
I  9  10  6, 25  10  t  e
5 4
e
II  9 10  6, 25  10   t  T   e 1,5610   6,910  
5 t T 6 t T
5 4
e
III  9 10  6, 25  10   t  2T   e 1,5610 
5 t 2 T 
6,910 
6 t 2 T 
5 4
e
IV  9  10  6, 25  10   t  3T   e 1,5610 
5 t 3T 
6,910 
6 t 3T 
5 4
e

169

By Vasfi Emre Ömürlü, Ph.D., 2005

Ex-5

Overall system response:

 tT y(t)  (F0 / T)  I 

 T  t  2T y(t)  (F0 / T)   I  II  
 
 
 2T  t  3T y(t)  (F0 / T)   I  II  III  

 3T  T   y(t)  (F0 / T)   I  II  III  IV  


170

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5

Final value teorem:

lim f (t )  lim sF ( s )
t  s 0

1,56105 t 6,9106 t
y(t)  9 10  6, 25  10  t  e
5 4
e
lim y(t)  9 10 5    1  1  
t 

1 1 1
Y(s)   2  2
2 s  60s  800 s
1 1 1 1
lim sY(s)   2   
s0
2 s  60s  800 s 0
lim y(t)  lim sY(s) 171
t  s0

By Vasfi Emre Ömürlü, Ph.D., 2005

 Ex-5

Initial value teorem:

f (0)  lim sF ( s )
s 
  
should  exist

1 1 1
Y(s)   2  2
2 s  60s  800 s
1 1 1 1
y(0 )  lim sY(s)   2   0
s 
2 s  60s  800 s 
172

By Vasfi Emre Ömürlü, Ph.D., 2005