DMX5403 - Control Systems

Engineering

All are Welcome !

Department of Mechanical Engineering,

Department of Electrical and Computer Engineering
 Academic Coordinators -
Ms. Hemali Pasqual (Dept. of Elect. & Comp.)
Dr. G. Anthonys (Dept. of Elect. & Comp.)
Mr. H.D. Nelaka Shayamal Priyanakara (Dept. of Mechanical Eng.)
Mr.A.C.G. Perera(Dept. of Mechanical Eng.)

Course Coordinators –
Mr. Hiran Wijesuriya (Dept. of Mechanical Eng.))
Email:[email protected]

Course Materials - Book 1, Book 2 & Book 3

Course Consists -
2 TMAs - TMA1, & Online quiz
2 CATs - CAT1 - OBT – Based on Book1
CAT2 – CBT Based on Book 1 & 2
Mini Project
Final Examination - Closed Book Type (Three hours)

2
DMX4543
 CA mark
CA mark = 1B(CAT)0.4 + 2B(CAT)0.3 + AVG(TMA,OQ)0.1 + AVG(LAB,MP)0.2
>= 40
&
LAB>=50
MP>=50
Eligibility criteria :
If [LAB] ≥ 50% AND [CA] ≥ 40%
Virtual Class :
You can access the moodle class from:
http://elearn.ou.ac.lk/
User name: S-ID Password-NIC

Very Important:
# Late submissions will not be accepted.
# Labs are compulsory.
3
DMX4543
In this course …
• Introduction to control systems
• Modeling of control systems
– Mathematical equations, Signal Flow Graphs,
Transfer function representation, Block diagram
representation, State Space Representation
• Time and Frequency domain analysis
– Transient analysis, Steady state analysis
• Stability of control systems
– Routh Hurwitz criteria, Root Locus plots
– Bode plots, Nyquist plots, Nichol's charts
• Design of control systems using classical
methods
– PI, PD, PID controllers etc.
• Use of MATLAB & Simulink in control systems
4
DMX4543
 References:

• B. C. Kuo: Automatic Control Systems, Prentice-Hall, Inc.

• D.K. Anand, R.B Zmood: Introduction to Control Systems,
Butterworth-Heinemann, 1995
• Richard C. Dorf, Robert H. Bishop: Modern Control
Systems, Prentice, 2001
• K. Ogatha, Modern Control Systems: Prentice Hall Inc.
• Norman S. Nise, Control Systems Engineering Addison-
Wesely Publishers etc.
• www.me.cmu.edu/ctms/controls/index.htm
• www.mathworks.com/products/control
• http://claymore.engineer.gvsu.edu/~jackh/books/model/

You may also use some other books/ web sites of your choice.
5
DMX4543
 Contents:
• Overview
• Introduction to Control Systems
• Mathematical Modelling and
Transfer Functions
• Transfer Function Calculations
• Block Diagrams
• Signal Flow Graphs
• State Space Representation
• Transient Analysis
• Steady state Analysis
6
DMX4543
 Overview

Classical Methods Classical Controllers

(Root locus, bode plot . . .) (P, PI, PID … controllers)

Model Stability ? To Stable

Nonlinear methods Intelligent controllers

(SS, Adaptive . . .) (Fuzzy, Neural, Expert . . .)

7
DMX4543
 Introduction
In a system there are three major stages
as follows:

Input(s) Process Output(s)

According to the Inputs/ Outputs any

systems can be categorized as follows:
SISO SIMO MISO MIMO
8
DMX4543
 Then it can be illustrated a simple control
system as follows:
Input Output
Controller Process

Open loop control system

What are these?

Input Output
Flow path
Controller Process

What is meant by Open loop?

9
DMX4543
 Further this can be illustrated as follows:
Input Output
Controller Actuator Process

Sensor

Closed loop control system

What are these?
Input Actuator Controller
Summing point Process Sensor Output

What is Forward path/ Feedback path?

10
DMX4543
 Further on feedback control systems

11
DMX4543
 Example

Suppose you need to walk to a particular

place.
- With your eyes open
- With your eyes closed

- Can you draw the block diagrams for

each case?

12
DMX4543
 •When eyes are open – closed loop control system
Input Output
Controller Actuator Process
Desired Brain Legs Walking
Actual
Destination
destination

Sensor
Eyes

•When eyes are closed – open loop control system

Input Controller
Actuator Process
Output

Desired Legs Walking

Brain Actual
Destination
destination

13
DMX4543
 • Note that a typical control system
includes the sensors, actuators, and the
control law.
– The sensors and actuators need not
always be physical devices.
– A good selection of the sensor and
actuator can greatly simplify the control
design process.
– The design of the control law given the
rest of the system (although we will need to
model the system).
14
DMX4543
 Why Use Control?
• Typically easy question to answer for
aerospace because many vehicles
(spacecraft, aircraft, rockets) and
aerospace processes (propulsion) need
to be controlled just to function.

• But there are also many stable systems

that simply require better performance
in some sense (e.g., faster, less
oscillatory), and we can use control to
modify this behavior.
15
DMX4543
 Feedback Control system example:

F-117 from U.S. Air Force

16
DMX4543
• Establish control objectives
– Qualitative : e.g. don’t use too much fuel
– Quantitative : e.g. settling time of step response
<3 sec
– Typically requires that you understand the
process (expected commands and
disturbances) and the overall goals
(bandwidths).
– Often requires that you have a strong
understanding of the physical dynamics of
the system so that you do not “fight” them
in inappropriate (i.e., inefficient) ways.

17
DMX4543
 • Select sensors & actuators
– What aspects of the system are to be
sensed and controlled?
– Consider sensor noise and linearity as key
discriminators.
– Cost, reliability, size, . . .

• Obtain model
– Analytic (FEM) or from measured data
– Evaluation model → reduce size/
complexity → Design model
– Accuracy? Error model?
18
DMX4543
 • Design controller
– Select technique (SISO, MIMO), (classical,
state-space)
– Choose parameters (ROT, optimization)

• Analyze closed-loop performance.

Meet objectives?
– Analysis, simulation, experimentation
– Yes → done, No → iterate . . .

19
DMX4543
 Example: Blimp Control
• Control objective
– Stabilization
– Red blimp tracks the motion of the green blimp

• Sensors
– GPS for positioning
– Compass for heading
– Gyros/GPS for roll altitude

20
DMX4543
 • Actuators
– electric motors (propellers) are very
nonlinear.
• Dynamics
– “rigid body” with strong apparent mass
effect.
– Roll modes.
• Modeling
– Analytic models with parameter
identification to determine “mass”.
• Disturbances
– wind
21
DMX4543
 22
DMX4543
 23
DMX4543
 As an example: Manual control system for regulating
the level of the fluid in a tank

Visual (Sensor)

Operator
(Controller)

Desired level
(Reference)

Valve (Actuator)
24
DMX4543
Desired Controller Actuator Process
level Actual
level
Operator Valve Tank

Visual

Sensor

Block diagram of the System

25
DMX4543
 How model a system?

Mathematical representation
-Transfer functions
- State space representation

Graphical representation
-Block diagrams
-Signal Flow Graph

26
DMX4543
 Mathematical Modelling and
Transfer Functions

output
t. f 
input
C(s) Output in Laplace domain
=
R(s) Input in Laplace domain
27
DMX4543
 Properties of the Transfer Function
1. Transfer function is defined only for a
linear time-invariant system. It is not
defined for non-linear systems.
2. The transfer function can be defined:
LaplaceTransformoutput variable
Transfer function 
LaplaceTransforminput variable

3. All initial conditions of the system are

set to zero.
4. The transfer function is independent
of the input of the system.
28
DMX4543
 Transfer Function Calculations

29
DMX4543
• Basic Building blocks – Mechanical systems

Impedance
Component Force - velocity Force - displacement
Z M (s)  F (s) / X (s)

Spring
x(t )

t
f (t )  K  v( )d f (t )  Kx (t ) K
f (t ) 0
K

Viscous damper

x(t )
dx(t ) fvs
f (t )  Fv v(t ) f (t )  f v
f (t ) dt

fv

Mass
x(t )
dv(t ) d 2 x(t ) Ms 2
f (t )  M f (t )  M
M f (t ) dt dt 2

30
DMX4543
 Eg:
Find the transfer function X(s) F(s) of the
following Spring-mass damper system.
K

M f (t )

c x(t )

Assume the mass travelling to the right.

31
DMX4543
 Applying Newton’s law:
2
d x(t ) dx(t )
M c  Kx(t )  f (t )
2 dt
dt
32
DMX4543
 Taking the Laplace transform,
assuming zero initial conditions:

Ms X ( s )  csX ( s )  KX ( s )  F ( s )
2

2
( Ms  cs  k ) X ( s )  F ( s )
X ( s) 1
G ( s)  
F ( s ) Ms  cs  K
2

33
DMX4543
 We can obtain the following equations:

For the spring, F (s)  KX (s)

For the viscous F (s)  CsX (s)

damper,

2
For the mass, F (s)  Ms X (s)

34
DMX4543
 Block Diagrams
Block diagram of a feedback control system
+
R E C
G

G = Direct transfer function.

H = Feedback transfer function.
GH = Open loop (or loop) transfer function.
C/R = Closed loop transfer function.
E/R = Actuating signal ratio. (or error ratio.)
B/R = Primary feedback ratio.
35
DMX4543
 Simplifying the Block diagrams

Eg. Reduce the following diagram.

G3

R G1 G4 G2 C

H1

H2

36
DMX4543
 Step 1:
Combine all cascade block.

G1 G4  G1G4

Step 2:
Combine all parallel blocks.

37
DMX4543
 Step 3: Eliminate all minor feedback loops

Step 4: Now we can draw the block diagram as:

G1G4
G1  G2
1  G1G4 H 1

H2

38
DMX4543
 Finally we get the following canonical form.

G1G4 G1  G2 
1  G1G4 H 1

H2

39
DMX4543
 Signal Flow Graphs
Signal Flow Graph from block diagrams

R(s) C(s) R(s) G(s) C(s)

G

H(s)

40
DMX4543
 Eg. Draw the signal flow graph of the
following block diagram.

41
DMX4543
Solution

42
DMX4543
 Mason’s Gain Formula Yout M k k
N
M  
Yin k 1 
Where,
y in  i p node variable. yout  o p node varible.
M = gain between and
N = total number of forward paths between &
Mk = gain of the kth forward path between &
  1   Li1   L j 2   Lk 3  ......
i j k

Lmr = gain product of the mth (m = i, j, k……)

possible combination of nontouching loops
43
DMX4543
 Eg. Determine the gain between y1 and y5 of
the following SFG using Mason’s gain formula.

Solution

1) The 3 forward paths between y1 and y5 are,

M 1  a12 a 23 a34 a 45 y1  y 2  y 3  y 4  y 5
M 2  a12 a 24 a 45 y1  y 2  y 4  y 5
M 3  a12 a 25 y1  y 2  y 5
44
DMX4543
2) There are 5 individual loops.
L 11  a 23 a 32 L 12  a 34 a 43 L 13  a 24 a 43 a 32
L 14  a 44 L 15  a 24 a 44 a 43 a 32

3) There is only 1 pair of nontouching loops

y 2  y3  y 2 & y 4  y 4
3

i.e. L11 & L44

The product of the gain of the 2 nontouching loops is L21  a 23 a32 a 44

4) All the loops are in touch with forward paths M1 and M2

 1  1 & 2  1
2 of the loops are not in touch with forward path M
y3  y 4  y3 & y4  y4
  3  1  a34 a 43  a 44
45
DMX4543
Substituting above,

  1  L11  L12  L13  L14  L15   L21

 1 - a 23 a32  a34 a 43  a 24 a32 a 43  a 44  a 24 a 44 a 43 a32   a 23 a32 a 44

Gain between y 5 & y1

y5 M 11  M 2  2  M 3  3

y1 

a12 a 23 a34 a 45  a12 a 24 a 45  a12 a 25 1  a34 a 43  a 44 


1  a 23 a32  a34 a 43  a 24 a32 a 43  a 44  a 24 a 44 a 43 a32   a 23 a32 a 44

46
DMX4543
 Transient Analysis
After we obtain a mathematical representation
of a subsystem, it is then analysed for its time-
domain response (transient and steady-state
responses) to see if these characteristics yield
the desired behaviour.

Response Characteristics
A control system provides an output or response
for a given input or stimulus.
Input - represents a desired response
Output - represents the actual response
47
DMX4543
 Consider the operation of an elevator system.

You want to use the elevator to go to the

4th floor from the ground floor.
When you push the 4th floor button,
the elevator rises to the 4th floor with
an acceleration, a speed and floor
levelling accuracy designed for
passenger comfort.

The elevator itself follows the displacement

described by the curve marked as elevator
response.
48
DMX4543
 49
DMX4543
 Poles, Zeros and System Response

Poles of a transfer function is:

The values of Laplace Transform variable
s, that cause the transfer function to
become infinite.

Zeros of a transfer function is:

The values of Laplace Transform variable
s, that cause the transfer function to
become zero.

50
DMX4543
 First order systems

51
DMX4543
 Time Constant –
We call 1 a as the time constant of the response.
Rise Time –
Rise time can be defined as the time taken for
the waveform to go from 0.1 to 0.9 of its final
value.

Settling Time -
Settling time is the time for the response to
reach and stay within 2% of its final value.

52
DMX4543
 Second order systems
Over damped response -
9 9
C (s)  
s s  9 s  9  ss  7.854 s  1.146 
2

7.854 t 1.146 t
The output, c(t )  1  0.171e  1.171e

53
DMX4543
 Under damped response -
9 9
The output response C(s)  =
s(s  2s  9)
2
s(s 1 j 8)
c(t )

s  1  j 8

54
DMX4543
 Un-damped response -
9 9
The transfer function, C ( s )  
ss  9 ss  3 j 2
2

output c(t )  K1  K 4 Cos3t   

55
DMX4543
 Critically Damped Response -
9 9
The transfer function, C ( s )  
s s  6 s  9  s s  3
2 2

56
DMX4543
 Under damped second order systems
The two parameters associated with
second order systems, and  n 
 - Damping ratio  n - Natural frequency
Transient Specifications
.

57
DMX4543
Note :
All these definitions are also valid for systems of higher order.

58
DMX4543
 This information can be used to
determine whether the nature
of the response do or do not
degrade the performance of the
system.

59
DMX4543
 Steady state Analysis
What is steady state error?
We can define the steady state error as the
difference between the input and the output
for a given test input as, t  

Any physical control system inherently suffers

from steady state errors in response to certain
types of inputs. These inputs can be classified
into 3 main types as given in Table below.

60
DMX4543
 Name Physical Time Laplace
interpretation Function Transform

Step constant
position 1 1/s

Ramp constant t 1/s2

velocity

Parabola constant
accelerator 1/t2 1/s3

61
DMX4543
 Type of a control system
Consider a unity feed back control system
with the following open loop transfer function:
K(Ta s  1 )(Tb s  1 ).....(Tm s  1 )
G s   N
S T1 s  1)(T2 s  1)......(T p s  1
N
The term S in the denominator represents
a pole of multiplicity N at the origin.
A system is called ‘Type 0’ when N = 0
‘Type 1’ when N = 1
‘Type 2’ when N = 2 etc.
Note that this is different from the ‘order’ of a system.
62
DMX4543
 Sensitivity of a control system
Sensitivity means the degree to which changes
in the system parameters affect system transfer
function and hence performance.
Assuming a function F and a parameter P;

63
DMX4543
 Eg. Consider the system

R(s) + Ea (s) K C(s)

s ( s a)
-

Calculate the sensitivity of the closed loop

transfer function to changes in the parameter a.

64
DMX4543
 Solution
K
Closed Loop Transfer Function 
s 2  as  K
a δT
sensitivity s T:a 
T δa
a  Ks
 . 2
 K  (s  as  K) 2
 2 
 s  as  K 
 as

s 2  as  K

This is a function of the value of s. The

increase in K will reduce the sensitivity of
the closed loop transfer unction for any
value of s to changes in the parameter a.
65
DMX4543
 MATLAB & Simulink
MATLAB is a high-performance
language for technical computing.
It integrates :
• Computation,
• Visualization, and
• Programming

66
DMX4543
 MATLAB : Text mode
x = -pi:.1:pi;
y = sin(x); 1
plot(x,y) 0.8

0.6

0.4

0.2

-0.2

-0.4

-0.6

-0.8

-1
-4 -3 -2 -1 0 1 2 3 4

67
DMX4543
MATLAB : GUI mode

68
DMX4543
 Simulink is a software package for
modeling, simulating, and analyzing
dynamic systems.
Eg: Automotive Suspension

69
DMX4543
 70
DMX4543
Running the
Simulation

71
DMX4543
 A contour plot with
gradient arrows shows the
temperature & heat flux.

The temperature
gradient is displayed
using 3-D plot tools.
www.mathworks.com

72
DMX4543
 Why MATLAB/ Simulink?
• Accuracy
• Extremely fast
• Can apply to Complex systems
• Algorithm development
• Modeling, simulation, and prototyping
• Data analysis, exploration, and
visualization
• Scientific and engineering graphics
73
DMX4543
 In next sessions…
• Analysis of systems
– Transient analysis
– Steady state analysis
• Stability of systems
– How it is check?
– Time domain & frequency domain
• Design of control systems
– Using classical Controllers

74
DMX4543
 Thank you.

75
DMX4543