System Modeling Coursework
Class 2: Mathematical Modeling of systems using
transfer function approach
P.R. VENKATESWARAN
Faculty, Instrumentation and Control Engineering,
Manipal Institute of Technology, Manipal
Karnataka 576 104 INDIA
Ph: 0820 2925154, 2925152
Fax: 0820 2571071
Email: [email protected], [email protected]
Blog: www.godsfavouritechild.wordpress.com
Web address: http://www.esnips.com/web/SystemModelingClassNotes
WARNING!
• I claim no originality in all these notes. These are the
compilation from various sources for the purpose of
delivering lectures. I humbly acknowledge the
wonderful help provided by the original sources in
this compilation.
• For best results, it is always suggested you read the
source material.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 2
Why mathematical model is required?
• It is required to understand, analyse and control the
system
• Fundamental physical laws of science and
engineering are used for modeling.
– Electrical Systems: Ohms, Kirchoffs and Lenz law
– Mechanical Systems: Thermodynamic and Newton’s law
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 3
Introduction to mathematical modeling
• The mathematical model is usually in the form of
differential equations. A differential equation can
describe relationship between input and output
• For a linear system, Laplace transform can be used
to find the solutions of the differential equations
• Using Laplace Transforms, we can represent the real
system using transfer functions.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 4
Introduction
• Deriving reasonable mathematical models is the most
important part of the entire analysis.
• The principle of causality is assumed throughout this course.
• A trade-off exists between simplicity and accuracy.
• Linear Systems.
– Superposition applies.
• Linear Time-Invariant (vs. Time-Varying) Systems.
• Differential equations with constant coefficients (vs. coefficients of
functions of time).
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 5
Transfer function approach
• The physical system, the input and the output
signals can be separated and easily visualized.
• Transfer function in the Laplace domain is that
relation which algebraically relates the input and
output of a control system, for the case when the
initial conditions are zero.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 6
Transfer function approach
Consider the linear time-invariant system defined by
the differential equation:
(n) ( n −1) . (n) ( n −1) .
a0 y + a1 y + ⋅⋅⋅ + an −1 y + an y = b0 x + b1 x + ⋅⋅⋅ + bn −1 x + bn x ( n ≥ m)
where x = input and y = output
L[ y ]
Transfer function = G ( s ) =
L[ x] zero initial conditions
Y ( s ) b0 s m + b1s m −1 + ⋅⋅⋅ + bm −1s + bm N ( s )
= = n −1
=
X ( s ) a0 s + a1s + ⋅⋅⋅ + an −1s + an D( s )
n
Order ( D( s )) = number of order of the system.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 7
Characteristics of transfer function
• It is mathematical model to express the relationship between the
output and the input variable.
• It is independent of the magnitude and nature of the input or driving
function.
• It includes the units. However it does not provide any information
concerning the physical structure.
• If it is known, the output or response can be studied for various
forms of inputs.
• If it is unknown, it may be established experimentally by introducing
known inputs and studying the output. Once established, it gives a full
description of the dynamic characteristics.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 8
Example: Impulse response function
Convolution Integral.
Y (s) = G (s) X (s)
t t
⇒ y (t ) = ∫ x(τ ) g (t − τ )dt = ∫ g (τ ) x(t − τ )dt
0 0
where g (t ) = x(t ) = 0 for t < 0.
Let x(t ) = δ (t ).
⇒ X (s) = 1
⇒ Y (s) = G (s)
L−1[G ( s )] = g (t ): Impulse-Response Function.
The transfer function and the impulse-response
function contain the same complete information
about the system dynamics.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 9
Laplace transform
• Laplace transforms provide a method for representing and analyzing
linear systems using algebraic methods.
• In systems that begin undeflected and at rest the Laplace can directly
replace the d/dt operator in differential equations. It is a superset of
the phasor representation in that it has both a complex part, for the
steady state response, but also a real part, representing the transient
part.
• As with the other representations the Laplace s is related to the rate of
change in the system.
D=s
s = σ + jω
(if the initial conditions/derivatives are all zero at t=0s)
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 10
Procedure to apply Laplace transform
• Convert the system transfer function, or differential equation, to
the s-domain by replacing ’D’ with ’s’. (Note: If any of the initial
conditions are non-zero these must be also be added.)
• Convert the input function(s) to the s-domain using the transform
tables.
• Algebraically combine the input and transfer function to find an
output function.
• Use partial fractions to reduce the output function to simpler
components.
• Convert the output equation back to the time-domain using the
tables.
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 11
Summary
• The understanding of the system is best obtained in linear
systems is from deriving the transfer function of the system.
• In order to make the analytical complexity easy to handle
the transformation of time domain to a different domain is
necessary. Laplace transformation helps in such
transformation. T
• he transfer function is obtained by identifying the input and
output variables from the system and deducing using
Laplace transforms
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 12
References
• http://en.wikipedia.org/wiki/Laplace_transform amongst
others…
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 13
And, before we break…
• The impossible is the untried
Thanks for listening…
July – December 2008 prv/System Modeling Coursework/MIT-Manipal 14
