ME7732:

Mechatronic Design and Automa?on

Basics of Control Engineering

Lecture 1
•  Ogata K, Pren%ce Hall, Modern Control Engineering
•  Richard C Dorf and Robert H Bishop, Modern Control Systems
•  William Bolton, Mechatronics - Electronic Control Systems in Mechanical
and Electrical Engineering


Dr Beenish Ayaz
[email protected]
 Brainstorming Ac?vity

1.  Have you ever tried to balance a

s?ck upright on your palm?

2.  How does the hea?ng system in

home ensure the desired
temperature?

3.  What is cruise control?

Overview

•  Introduc?on
•  System Modelling & Linearisa?on
•  First and Second Order Systems
•  System Stability
Why Study Control?
Control Applica?ons are numerous

Ø  Most modern engineering systems use CONTROL TECHNOLOGY

Ø  Automa?c control of machines and system Speed, Temperature, Trajectory

Control System Engineering

Focuses on the modeling, analysing and

designing of a diverse range of dynamic
systems and controllers that will cause these
systems to behave in a desired manner.
Control Design Problem

SYSTEM
Actuators Sensors

Plant

Design
Specifications
Modelling

Analysis and
DESIGN
Simulation

Implementation
Basic Terms & Concepts
System : A combina?on of components ac?ng together to perform a
specific objec?ve





Control System – An interconnec?on of components forming a system
configura?on that will provide a desired response.
Modeling : A procedure to obtain a model describing important
characteris?cs of system
Analysis : Inves?ga?on of performance of system, whose model is known,
under specified condi?ons
Examples of Modern Control Systems
Control System Types

1.  Open Loop Control System

2.  Closed Loop/ Feedback Control System.

Hea?ng Element Example

Temperature Control System

1.  Manual Hea?ng Element
- Open Loop Control System
2.  Hea?ng Element with Timer

3.  Hea?ng Element with Sensor - Closed Loop/ Feedback

Control System
Open Loop Control System

•  A system in which the control ac?on is totally independent of

the output.
•  It u?lizes an actua?ng device to control the process directly
without using feedback.

Single-Input, Single-Output (SISO) Open Loop System

Open Loop Control System
•  Fixed control ac?on
•  Can not adjust to unexpected changes/disturbances from the
system environment

Quiz
1.  Is Lawn Sprinkler an open loop control system?
2.  What if it starts raining (weather), will it stop sprinkling?
3.  How can you improve this system?

Lawn Sprinkler
 Closed Loop/Feedback Control System
•  In closed loop control system the output is measured
con?nuously and is fed back to the input.
•  The presence of feedback compensates for the disturbance
and improves the accuracy of the system.

Single-Input, Single-Output (SISO) Closed Loop System

Mul?-Input, Mul?-Output (MIMO) Closed Loop System

Closed Loop Control System - Example

The driver uses the difference

between the actual and the
desired direc?on of travel to
generate a controlled adjustment
of the steering wheel.

Automobile steering control system

15
Components of a Feedback Control System

Controller – Device/s that compute and generate control signals/ac?ons

Actuator – Device/s that perform control ac?on
Plant/Process – System to be controlled
Sensor – Device/s that measure the output
16
Linear and Time Invariant System (LTI)

Time-invariant System
•  The features of system never ever change.
•  Usually a good approxima?on for most engineering applica?on.
•  A good star?ng point to obtain main features of system.
•  Rela?vely easy to analyze

Linear System
•  Equa?ons describing system are linear
•  Principle of superposi?on
- For a linear system with excita?on [x1(t) and x2(t) ] and a response [y1(t) and y2(t)]
- x1(t)+ x2(t) = y1(t) + y2(t) this is called the property of superposi+on
•  Property of Homogeneity
- If x is input and y is output then β x = β y where β is constant.
Quiz
1.  What is a Control System?
2.  What is the difference between open-loop and closed-loop control
system?
3.  What is the purpose of Controller in control system?
4.  What does Actuator do?
5.  What is Plant/Process in a control system block diagram?
6.  Which block measures the output to feed back?
7.  What is a linear system?

Inverted Pendulum
hkps://www.youtube.com/watch?v=XWhGjxdug0o
0:40 – 2:00
18
System Modelling
Modelling
Deriving differen?al equa?ons for systems
•  Physical mathema?cal modelling is the process of analysing
a physical system to determine equa?ons which represent
its behaviour.
•  Ordinary differen?al equa?ons (ODEs) are used to represent
dynamic (i.e. ?me-varying) behaviour.
•  The process begins with deciding what is the system input,
and what is the output

20
System Equa?ons

•  Ooen a physical system can be modelled by an nth

order ODE of the form:

•  Some?mes, deriva?ves of the input variable are also

required:

21
Transfer func?on

X(s) Y(s)
G(s) Y(s)=G(s).X(s)

22
Laplace Transforms
Laplace Transform Table

24
System Modelling Using Laplace Transform

Mechanical systems can

generally be modelled as
spring, damper and mass.




Interdisciplinary System Nature

•  Models are the same regardless of the physical domain of interest


= = =

y = Charge y = Transl. displ. y = Angular displ. y = Volume
u = Voltage u = Force u = Torque u = Pressure


u 𝑎​𝑦 +𝑏​ 𝑦  +𝑐𝑦=𝑢(𝑡) y
•  We only need to understand how to analyze one model, but the
results are applicable for four seemingly different types of physical
systems!
System Modelling Using Laplace Transform

Spring with Damper with Mass M

s?ffness K damping
coefficient B

y
y
y F
F F
Viscous Force F = bv Newton’s law F = ma
Hooke’s law F = ky
dy d2y
F = ky F = b = by! F = m 2 = m!!
y
dt dt
F(s) = kY (s) F(s) = bsY (s) F(s) = ms 2Y (s)
System Modelling Using Laplace Transform
y(t)
K
f (t)
M

B
F = ma
Ky(t)
f (t) M!!
y(t)

By! (t)
−Ky(t) − By! (t) + f (t) = M!!
y(t)
System Modelling Using Laplace Transform

−Ky(t) − By! (t) + f (t) = M!!

y(t)

f (t) = M!!
y(t) + Ky(t) + By! (t)

Apply Laplace

F(s) = ( Ms 2 + Bs + K ) Y (s)

Y (s) 1
= 2
F(s) Ms + Bs + K
Example
Y (s) 1
=
Fi ( s ) Ms 2 + Bs + K

“ii. A 10N load is suddenly added to the trailer. Derive

the equa%on showing how y(t) varies aOer the addi%on
of the load.”

10
•  Input is 10 unit step input à Fi ( s ) =
s
Y (s) 1
= 10
10 2
Ms + Bs + K Y (s) =
s
(
s Ms 2 + Bs + K )
Finding inverse Laplace Transform of
this will give our response
Example contd.
f(t) F(s)
10 K = 140Nm-1
Y (s) = B = 90Nms-1
e-at
s ( Ms 2 + Bs + K )
M = 10kg A A/s
10 1 1
Y (s) = = =
10s ( s 2 + 9s +14) s ( s 2 + 2s + 7s +14) s ( s(s + 2) + 7(s + 2))

1 a b c 1
a=
Y (s) = = − + 14
s(s + 2)(s + 7) s (s + 2) (s + 7) 1
b=−
10
1 1 1 1
Y (s) = = − + c=
1
s(s + 2)(s + 7) 14s 10(s + 2) 35(s + 7) 35

1 1 1
y (t ) = − e −2 t + e −7t
14 10 35 Par?al Frac?on Calculator
hkps://www.wolframalpha.com/
widgets/view.jsp?
id=ec4a062bb304f88c2ba0b631
d7acabbc
So in Summary
y! (t ) + 2 y (t ) = f (t )
Taking the Laplace Transform

sY (s) + 2Y (s) = F(s)

Y (s) 1
=
F(s) s + 2
Taking inverse Laplace Transform
Laplace Transform Calculator
y(t) = e−2t f (t) hkps://www.emathhelp.net/
calculators/differen?al-
equa?ons/inverse-laplace-
From which we can now find solu?ons!! transform-calculator/
Proper?es of Laplace Transforms

o Linearity L [ af (t) + bg(t)] = aF(s) + bG(s)

⎡ d n f (t) ⎤ n
o Deriva?ves L⎢ n ⎥
= s F(s)
⎣ dt ⎦

o Integral ⎡ ∞ ⎤ F(s)
L ⎢ ∫ f (t)dt ⎥ =
⎣ 0 ⎦ s

33
Linearisa?on of dynamic equa?ons-Exp 1

For small values of y, sin y(t) ~ y(t) can be assumed,

Corresponding u at the steady operating point is 0. Ignoring higher

order terms:
Linearisa?on of dynamic equa?ons-Ex.3
First and Second Order Systems
System Classifica?on

Systems can be classified by the order of their transfer

func?on
E.g. Y (s)
= As + B = f ( y! (t ), y (t ))
R( s)
First Order System

Y (s)
= As 2 + Bs + C = f ( !y!(t ), y! (t ), y (t ))
R( s)
Second Order System
Test Input Signals

Step Ramp Parabolic

u(t) r(t) p(t)
First Order Systems

•  First order systems have transfer func?ons in the

form:
K
TF =
τs + 1

•  K = System gain
•  τ
= Time constant
Step Response - First
Order System

Step Response is the ?me response

of a system for step signal
 First Order System

1
T2 (s) =
(τ s +1)

⎛ 1 ⎞ a
τ
T2 (s) = ⎜ ⎟=
⎝s+ τ ⎠ s+a
1

t
1 τ⎤
⎡ −
T2 (t ) = ⎢e ⎥
τ⎣ ⎦

41
Step Response for different values of
Time Constant “ τ ”
Step Response
From: U(1)
1

K
0.9

G (s) = 0.8
τ =3
(τs + 1) 0.7
Amplitude
0.6
τ =1
τ = 0. 5
To: Y(1)

0.5

= 0. 2
0.4
τ
0.3

Increasing τ 0.2
increases ?me to
0.1
achieve steady
state (ss) 0
0 1 2 3 4 5 6 7 8 9 10

Time (sec.)

τ is changing, K=1, Unit step input

Step Response for different values of
System Gain “ K ”
Step Response
From: U(1)
2

K
G (s) =
1.8

(τs + 1) 1.6

1.4

1.2

K =2
Amplitude

To: Y(1)
1

K = 1 .5 0.8

0.6
K = 1 .2 0.4

K =1 0.2

0
0 1 2 3 4 5 6 7 8 9 10
Increasing K Time (sec.)
increases SS value
τ is consant, K varying, Unit step input
 Input u(t)

First Order Response Features ro

Time

STEP RESPONSE (step of size ro): dy

τ + y (t ) = Ku (t )
dt
Steady state value
yss=Kro
Steady state value of output:
the value of the output reaches a
yτ= 0.63 *Kro constant level

First order response

yss = K ro

Value of output at the Time

constant:
Output: y(t)
yτ= 0.63 Kro.
τ: time constant
Time
τ is the time taken for the output to
0 rise to 63% of the change in
0 10 20 30 40 50 60 output.

The system parameters do not change for different input signals.

Summary of Variables

τ = System Time Constant K = System Gain

•  Affects transient behaviour •  Affects steady state value
of system •  Mul?plies value of input
•  Increasing τ increases ?me to signal
achieve steady state •  Increasing K increases SS
•  Defined as amount of ?me value
for system to achieve 63% of
Steady State(SS) value
General Second Order System
Transfer func?on of a general second order system

Poles of the system

Case I – Overdamped System
i) If ζ>1 we have overdamped response and two different real poles :

Case II – Cri?cally damped system
ii) In the case of ζ=1 the system is cri?cally damped and we have two
repeated poles

Case III – Underdamped System
iii) If 0<ζ<1 we have underdamped response and two complex poles :

Unit Step Response

Y(s) =

The inverse Laplace transform of Y(s):

y(t) =
Step Response
y(t)

ωnt

•  ζ < 1 response osillatory with overshoot – underdamped system

•  ζ = 1 no overshoot (cri?cal damping) – cri?cally damped system
•  ζ > 1 no overshoot, but slow rise ?me – overdamped system
Second-order System : Damping Coefficient ζ

Underdamped System

52
Second-order System : Damping Coefficient ζ

53
Second-order System : Damping Coefficient ζ

Undamped System

V”

54
Performance of a Second-order System

Response of a second-order system for Unit Step Input

55
Performance of a Second-order System

Unit step response

ζ = 0.2

56
Performance of a Second-order System

57
Performance of a Second-order System

4 π
Ts = 4τ = Tp =
ζω n ωn 1 − ζ 2
−ζπ / 1−ζ 2 −ζπ / 1−ζ 2
M pt = 1+ e M pt % = 100e

2.16ζ + 0.60 1.8

Tr1 = ≈
ωn ωn

Formulae for control design

58
Example – In Matlab T ( s) =
25
( s 2 + 3s + 25)

59
Pole – Zero and System Stability
Poles and Zeros

•  The transfer func?on of any system consists of poles

and zeros
•  Poles are roots of denominator of transfer func?on
•  Zeros are roots of numerator of transfer func?on
•  Any real system must have more poles than zeros
•  If there are more zeros than poles, system is virtual
Poles and Zeros

Zeros

Poles
Stability

•  A system is stable if all Poles lie on the LHP

•  Stability is the characteris?cs of the system,
and not influenced by input(s)
•  Reasonable stability margins are required for
robustness
Poles
For a unit step input, the response Y (s)

65
Real Poles
Im

X X X X

Impulse − at Re

Response: Ke X
-3
X
-1 0
X X
+1

Time (s)
Complex Poles

s1,2 = −σ ∓ jω d

s1,2 = −ζω n ∓ jω n 1− ζ 2
Complex Poles
Im

X X X X

Impulse Re
X X X X
Response: -3 -1 0 +1

68
leo half right half
Pole Posi?ons v Time Response

70
Performance Specifica?ons

•  Steady State Performance

•  Transient Performance
•  Frequency Domain Performance
Transient Performance – 2nd Order

1.4
Step Response
Overshoot
tP
1.2
MP
1
90%

0.8
Amplitude

98%

0.6 Rise Time

0.4

Peak Time 2% Settling Time

0.2
10% tr
Sekling
0
tP tS Time
Time (sec)
Performance in s-plane
Overshoot

Sekling Time

Rise Time
Natural Frequency (ωn) and Damping ra?o (ξ)

s = −4 ± 3 j

a b

tan ϕ = b/a

ωn
b
ϕ
a

ωn2
Frequency Domain Performance
10
Resonance
Peak
5
1
M M=
0
2ζ 1 − ζ 2
dB amplitude

-3

-5

Resonance ζ = 0 .3
Frequency
-10

Bandwidth
ωr =-15ωn 1 − 2ζ 2
ω r ωn ω b
-20
-1 0 1
10 10 10

frequency
 System Zeros

•  Zeros can occur due to internal pathways between synergis?c

system poles
•  These pathways are evident through the mathema?cs

•  e.g.:
 System Zeros

As zero approaches origin there will be increase in overshoot

and decrease in Tp
 Steady State Gain

•  Amplifica?on of input signal by natural physical characteris?cs

of system

K ss = G ( s )
s →0

Example: ( s + 3)
G (s) =
( s + 4)( s + 5)

(0 + 3) 3
K ss = G ( s ) = = = 0.15
s →0 (0 + 4)(0 + 5) 20
References

1.  Dorf, R. and Bishop, R., 2017. Modern Control Systems.

2.  Ogata K, Pren%ce Hall, Modern Control Engineering
3.  William Bolton, Mechatronics - Electronic Control Systems in Mechanical and Electrical Engineering
4.  Mathworks.com. 2020. Mathworks - Makers Of MATLAB And Simulink. [online] Available at:
<hkp://mathworks.com/> [Accessed 2 July 2020].

79