ECE 3151 - Spring 2010

Mathematical Models of Systems

Part a

Prof. Marian S. Stachowicz

Laboratory for Intelligent Systems
ECE Department, University of Minnesota Duluth

January 26 - February 16, 2010

 Chapter 2: Objectives

• We use quantitative mathematical models of

physical systems to design and analyze
control systems.

• The dynamic behavior is generally described

by ordinary differential equations

Control Systems 2
• We will consider a wide range of systems,
including mechanical, hydraulic, and electrical.

• Since most physical systems are nonlinear, we

will discuss linearization approximations, which
allow us to use Laplace transform methods.

Control Systems 3
• The transfer function blocks can be organized
into block diagrams or signal-flow graphs to
graphically depict the interconnections.

• We will then proceed to obtain the input–

output relationship for components and
subsystems in the form of transfer functions.

Control Systems 4
• We conclude the chapter by developing transfer
function models for the various components of the
Sequential Design Example: Disk Drive Read
System.

Control Systems 5
 References for reading

1. R.C. Dorf and R.H. Bishop, Modern Control Systems,

11th Edition, Prentice Hall, 2008,
Chapter 2.1 - 2.11

2. J.J. DiStefano, A. R. Stubberud, I. J. Williams,

Feeedback and Control Systems, Schaum's Outline
Series, McGraw-Hill, Inc., 1990
Chapters 3, 4, 8

Control Systems 6
 Mathematical Models of Systems

• Review the Laplace transform,

• Learn how to find a mathematical model,

called a transfer function, for LTI
electrical, mechanical, and
electromechanical systems,

• Learn how to linearize a nonlinear system

in order to find the transfer function.
Control Systems 7
 Outline

• 2.1 Introduction
• 2.2 Differential Equations of Physical Systems
• 2.3 Linear Approximations of Physical
Systems
• 2.4 The Laplace Transform
• 2.5 The Transfer Function of Linear Systems
• 2.6 Block Diagram Models
• 2.7 Signal-Flow Graph Models
Control Systems 8
 2.1 Introduction

• To understand and control complex

systems we must obtain quantitative
mathematical models of system.

Control Systems 9
• A model is a representation of the
process or a system existing in reality
or planned for realization which
expresses the essential attributes of a
process or a system in a useful form.
Norbert Wiener, 1945

Control Systems 10
Control Systems 11
Approach to dynamic systems

• 1. Define the system and its components.

• 2. Formulate the MM and list the necessary
assumptions.
• 3. Write the differential equations describing the
model.
• 4. Solve the equations for the desired output
variables.
• 5. Examine the solutions and the assumptions.
• 6. If necessary, reanalyze or redesign the
system.

Control Systems 12
 2.2 Differential Equations of
Physical Systems

The differential equations describing the

dynamic performance of a physical
system are obtained by utilizing the
physical laws of the process.

Control Systems 13
This approach applies equally well to;
• Mechanical,
• Electrical,
• Fluid,
• Thermodynamic systems.

Control Systems 14
 Physical laws

The physical laws define relationships

between fundamental quantities and are
usually represented by equations.

Control Systems 15
 A differential equation

A differential equation is any algebraic or

transcendental equality which involves
either differentials or derivatives [3].

Control Systems 16
 Ex. 3.1 from [2]

• The scalar version of Newton’s second

law states that, if a force of magnitude f is
applied to a mass M units, the acceleration
a of the mass is related to f by equation

f=Ma f = M (dv/dt)

The velocity v = v(t) and force f = f(t) are dependent

variable, the time t is the independent variable.
Control Systems 17
 Ex. 3.2 from [2]

• Ohm’s law states that, if a voltage of

magnitude v is applied across a resistor of R
units, the current i through the resistor is
related to v by the equation

v=Ri v = R (dq/dt)

The charge q = q(t) and the voltage v = v(t) are dependent

variable, the time t is the independent variable.
Control Systems 18
 A linear equation

• A linear equation is an equation

consisting of sum of linear terms.

• A linear term is one which is first degree

in dependent variables and their
derivatives.

Control Systems 19
• If any term of differential equation
contains higher power, products, or
transcendental functions of the dependent
variables, it is nonlinear.

dy 3 dy
( ) , u , sin u
dt dt

Control Systems 20
 A time-invariant equation

• is an equation in which none of the terms

depends explicitly on the independent
variable time.

Control Systems 21
 A time-variable equation

• is an equation in which one or more terms

depend explicitly on the independent
variable time.

Control Systems 22
 Causality

• A system in which time is the independent

variable is called causal if the output
depends only on present and past values
of the input.
That is, if y(t) is the output, then y(t)
depends only on the x(td) for values of
td ≤ t.

Control Systems 23
 Are of these systems causal ?

• y(t) = u(t +T), T >0 prediction

The output depends only on the input T sec in the future.
It is not causal.
• y(t) = u(t - T), T >0 a time delay T
The output depends only on the input T sec in the past.
It is causal.

Control Systems 24
 Why are all transcendental
functions not of first degree?
• Transcendental functions, such as the logarithmic,
trigonometric, and hyperbolic functions and their
corresponding inverses, are not of first degree because they
are defined by as infinite series.
• For example,
 2 n1 3 5
sin x   (1)
x x x
n1
 x    ...
n1 (2n 1)! 3! 5!

Control Systems 25
 LTI Control Systems

Linear Time Invariant Control Systems

Control Systems 26
Symbols and units

Control Systems 27
Control Systems 28
Inductive storage

Electrical inductance

Translational spring

Rotational spring

Fluid inertia

Control Systems 29
Capacitive storage

Electrical capacitance

Translational mass

Rotational mass

Control Systems 30
Energy dissipators

Electrical resistance

Translational damper

Rotational damper

Control Systems 31
Energy dissipators

Fluid resistance

Thermal resistance

Control Systems 32
 An automobile shock absorber

Spring-mass-damper Free-body diagram

Control Systems 33
Differential equations of physical
systems

Control Systems 34
 Equation 2.1

d 2 y(t) dy(t)
M 2
b  ky(t)  r(t)
dt dt



Control Systems 35
 Newton’s second law of motion
2
d y(t ) dy(t )
M 2
b  ky(t )  r (t )
dt dt
k is a spring constant
b is the friction constant

y(t)= K1 exp(-1t) sin(1t+1)

Control Systems 36
Control Systems 37
©Oxford University Press 2001
Control Systems 38
©Oxford University Press 2001
v(t)/R + C dv(t)/dt +1/L v(t) dt = r(t)

Control Systems 39
 Kirchhoff’s current law

v(t)/R + C dv(t)/dt +1/L v(t) dt = r(t)

v(t)= K2 exp(-2t) cos(2t+2)

Control Systems 40
Voltage response for underdamped
RLC circuit.

Control Systems 41
Solutions of the differential equations

• y(t)= K1 exp(-1t) sin(1t+1)

• v(t)= K2 exp(-2t) cos(2t+2)

Control Systems 42
 The close similarity

• M d2y(t)/dt2 + b dy(t)/dt + k y(t) = r(t)

for v(t) = dy(t)/dt

• M dv(t)/dt + b v(t) + k v(t) = r(t)

C dv(t)/dt + v(t)/R +1/L v(t) dt = r(t)

Control Systems 43
 Analogous variables

Voltage-velocity analogy:

voltage v(t) and velocity v(t) are equivalent

variables

Current-velocity analogy:

current i(t) and velocity v(t) are equivalent

variables
Control Systems 44
Control Systems 45
©Oxford University Press 2001
2.3 Linear Approximations of
Physical Systems
 ∆f = m ∆y

Control Systems 47
 Taylor Series Review

y(t) = g(x(t))

y = y0 + m (x - x0)

The linear equation: ∆y = m ∆x

Control Systems 48
 Taylor Series for Multi-Variable
Functions

   
f i1  I0  f (x g  X 0 ) 
f (i1, x g )  f (I0 , X 0 )     ......
i1 i1  I 0 
 1!  x g

x g  X 0  1! 
 xg  X 0   i1  I 0 


Control Systems 49
 Pendulum oscillator model

T= M g L sin  The nonlinear relation

Control Systems 50
 The linear approximation

Nonlinear relation between T and Q:

T= M g L sin 

The linear approximation relation between T and Q:

Ta= M g L 

• The approximation is reasonably accurate for

- /4 ≤  ≤  /4
Control Systems 51
 Approximation error

Calculate:

T - Ta for  = /6

Control Systems 52
 E 2.2 Thermistor

• A thermistor has a response to temperature

represented by
R = R0 e- 0.1T

R0= 10 [k]
T [C]
Find the linear model for thermistor at 20C
and small range of variation of temperature.
Control Systems 53
Thermistor

Control Systems 54
E2.3 Spring

Control Systems 55
Control Systems 56
Control Systems 57
Control Systems 58
Control Systems 59
Control Systems 60
Control Systems 61
Control Systems 62
Plot[{r2, (-1 + (1 + 4 r)/2}, {r, 0, 4}]

Control Systems 63
Control Systems 64
 Transforms

• The term transform refers to a mathematical

operation that takes a given function and returns a
new function.

• The transformation is often done by means of an

integral formula.

• Commonly used transforms are named after Laplace

and Fourier.

Control Systems 65
• Transforms are frequently used to change
a complicated problem into a simpler one.

• The simpler problem is then solved,

usually using elementary algebraic means.

• The solution to the simpler problem is

taken over to the original problem using
the inverse transform.

Control Systems 66
Control Systems 67
 Laplace transform

• Laplace transform can significantly reduce the

effort required to solve linear differential equations.

• A major benefit is that this transformation convert

differential equations to algebraic equations, which
can simplify the mathematical manipulations
required to obtain a solution.

Control Systems 68
Control Systems 69
 General solution procedure:

• Step 1. Take the Laplace transform of both sides of the

differential equation.

• Step 2. Solve for Y(s)

If the expression for Y(s) does not appear in Laplace Transform
Table
• Step 3a. Factor the characteristic equation polynomial.
• Step 3b. Perform the partial fraction expansion.

• Step 4. Use the inverse Laplace transform relations to find y(t).

Control Systems 70
 Example 2.2

d2y dy
2
 4  3 y(t )  2r (t )
dt dt

dy
y(0) 1, (0)  0, r(t) 1, t  0
dt

[s 2Y (s)  sy (0)]  4[sY (s)  y(0)]  3Y (s)  2R(s)

Control Systems 71
 Example 2.2
( s  4) 2
Y ( s)  2 
( s  4 s  3) s ( s 2  4 s  3)

q(s)  s 2  4s  3  (s  1)( s  3)  0

 3 / 2  1/ 2    1 1/ 3  2 / 3
Y ( s)          Y1 (s)  Y2 ( s)  Y3 (s)
 (s  1) (s  3)   ( s  1) (s  3)  s

Control Systems 72
 Example 2.2

 3  t  1  3t   t 1  3t  2
y (t )   e  e    1e  e  
2 2   3  3

2
lim t  y(t) 
3


Control Systems 73
 In[5]:=
Plot[3/2 Exp[-t]-1/2 Exp[-3t]-Exp[-t]+1/3 Exp[-
3t]+2/3,{t,0,6}]

Control Systems 74
 Example by Mathematica

p[s]:=1/(s+2s2+s3)

Factor[p[s]]

(1/(s (1 + s)2))

Control Systems 75
 Disadvantage:

• The solution of the differential equation involves use of

Laplace transforms as an intermediate step.

• Any change in the initial conditions or in the forcing

function requires that the complete solution be redeliver.

Control Systems 76
 The transfer function - a modified
approach.

• The transfer function is an algebraic expression for

the dynamic relation between input and output of the
process model.

• It is defined so as to be independent of initial

conditions and of the particular choice of forcing
function.

Control Systems 77
 G(s)

• To obtain the transfer function G(s) of the

LTI system, we take the Laplace transform
on both sides of the equation, and assume
zero initial conditions.

Control Systems 78
 Properties of the G(s)

• The G(s) is defined only for a LTI system.

• All initial conditions of the system are set to zero.
• The G(s) is independent of the input of the system.
• The G(s) of a continuous-data system is expressed
only as a function of the complex variable s.
• For discrete-data systems modeled by difference
equations, the transfer function is a function of z
when the z-transform is used.

Control Systems 79
• A transfer function can be derived only
for a LTI differential equation model.

Control Systems 80
 A transfer function

• A transfer function of the LTI system is

defined as a ratio of the Laplace transform
of the output variable to the Laplace
transform of the input variable, with all
initial conditions assumed to be zero.

Control Systems 81
 An automobile shock absorber

Spring-mass-damper Free-body diagram

Control Systems 82
 The automobile shock absorber

d 2 y(t ) dy(t )
M 2
b  ky(t )  r (t )
dt dt

Output Y(s) 1
 G(s)  
Input R(s) Ms 2  bs  k


Control Systems 83
 Transfer function of the RC network

V1(s) = (R + 1/Cs) I(s)

V2(s) = I(s) 1/Cs

G(s) = V2(s)/V1(s) = 1/(RC s + 1) = 1/T/(s + 1/T)

Control Systems 84
The transfer function of the RC network is obtained by writing
the Kirchhoff voltage equation.
The circuit is a voltage divider, where

V2(s)/V1(s) = Z2(s)/(Z1(s) + Z2(s)),

where Z1(s)= R and Z2 = 1/Cs
The single pole s = -1/T

Control Systems 85
• A transfer function of LTI system is
defined as the Laplace transform of the
impulse response, with initial conditions
set to zero.

Control Systems 86
 Input-Output description

• A transfer function is an input-output

description of the behavior of a system.

• Thus the transfer function description

does not include any information
concerning the internal structure of the
system.

Control Systems 87
 Control and politicians

• Control is a sequence of decisions aimed

at the attainment of specified objectives
in an environment of uncertainty and
presence of disturbances.

Control Systems 88
 Summary

1. The differential equations describing the dynamic

performance of physical systems were utilized to construct a
mathematical model. The physical systems included
mechanical, electrical, fluid, and thermodynamic systems.

2. A linear approximation using a Taylor series expansion

about the operating point was utilized to obtain a small-signal
linear approximation for nonlinear control control
components.

Control Systems 89
 Summary

3. For linear systems we apply the Laplace transformation

and its related input-output relationship given by the
transfer function.

4. The transfer function allows to determine the response

of the system to various input signals.

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

Control Systems 90
Control Systems 91
Control Systems 92
©Oxford University Press 2001
Control Systems 93
Control Systems 94
Control Systems 95
Control Systems 96
Control Systems 97
Study for Test 1: 4.15, 4.16,
4.17,4.22, 423, 4.24, 4.25, and
4.38 from Di Stefano…

Control Systems 98