Chapter 3 Modeling of Physical Systems

3.1 Foundation of Modeling

3.2 Method in Dynamic Modeling

3.3 Classification of Dynamic Model

3.4 Mathematical Representation
3.1 Foundation of Modeling (1)
Role & Use of Model:

Eykhoff
A representation of the essential aspects of an existing
object (system) which presents knowledge of that object
(system) in a usable form.

Websters dictionary
A description or analogy used to help visualize something
(as a system) that cant be directly observed.
3.1 Foundation of Modeling (2)
Modeling is an art rather than a technique
Behavior Objective
Representation
Optimal Model
3.1 Foundation of Modeling (3)
Principles of Modeling:


Principle of resemblance error tolerance

Principle of parsimony as simple as possible

Principle of objective use of information

3.1 Foundation of Modeling (4)
Method of Modeling:


Real
System
Model of Real
System
Information
Utilization
Experimental
Testing
Model
Building
System
Realization
Problem
Solving
Solution
Interpretation
Information
Processing
Behavior Representation
Objective
3.1 Foundation of Modeling (5)
Types of Model:

Conceptual model

Analog model

Graphical model

Solid model

Physical model

Mathematical model
3.2 Method in Dynamic Modeling (1)
Method in Dynamic Modeling:
Actual
System
Physical
Modeling
Mathematical
Modeling
Dynamic
Prediction & Analysis
Error
Acceptance
Test
No
Yes
Dynamic
Behavior
3.2 Method in Dynamic Modeling (2)
Physical Model:
An idealized physical system which resembles an actual system in its
salient features but which is more amenable to system analysis and synthesis.

Idealization Techniques:
Neglect small effect
Independent environment
Lumped characteristics
Linear relationships
Constant parameters
Neglect uncertainty and noise

3.2 Method in Dynamic Modeling (3)
Ex:
particle model
Newtonian particle
Fluid particle
Ideal gas
Photon

field model
Gravitational field
Flow field
Electromagnetic field
3.2 Method in Dynamic Modeling (4)
Mathematical Model:
The description of object behavior by means of suitably chosen
mathematical realizations.

Idealization:
Continuity
Directionality
Uniformity
Additivity
Constancy
Certainty

3.2 Method in Dynamic Modeling (5)
Model Verification:
Prediction error and acceptance criteria

Error Qualitative
Quantitative

Qualitative error phase plane portrait
Quantitative error e(t) (Well-controlled I.C.)



}
=
T
0
2
r m r m
dt (t) e
T
1
: RMS , ) t ( e : Ex
(t) e of function Norm : (t) e
state real : x , state model : x , ) t ( x ) t ( x ) t ( e



3.3 Classification of Dynamic Model (1)
Model Assumptions in this course:
Deterministic, Lumped, Linear, Time-invariant (Constant coeff.), and Continuous.

Dynamic Model
Determinstic Chaotic Stochastic
Lumped parameters
(ODE)
Distributed parameters
(PDE)
Linear
Non-linear
Costant coefficient
Variable coefficient
Discrete time Continuous time
3.3 Classification of Dynamic Model (2)
Various Physical and Mathematical Models of Real Pendulum:
J oint with clearance
Rod
g
m
mg
u
2
L
A
L
3.3 Classification of Dynamic Model (3)
Distributed and Lumped Model:
Distributed Model Lumped Model
m
mg
u
F
r
F
u
l
0 sin g
1
e. . i
sin mg m
F m
= u + u
u = u
= u
u
l


l

l
A
0 sin g
L ) 3 / 2 (
1
e. . i
)
2
L
( sin mg ) mL
3
1
(
M I
mL
3
1
I
2
A A
2
A
= u + u
u = u
= u
=

3.3 Classification of Dynamic Model (4)
Linear and Non-linear Model:
Non-linear Model Linear Model
0 sin
g
= u + u
l

t = u u
= u t + u
u t ~ u
t = u
= u + u
u ~ u
= u

u
+
u
u = u

g

g
e. . i
0 ) (
g

sin
point operating
pendulum Inv erted (2)
0
g

sin
0 point operating
pendulum Regular ) 1 (
.......
5 3
sin
5 3
l l

l

l

m
m
g
u
u sin
0 t t 2
3.3 Classification of Dynamic Model (5)
Time-invariant and Time-varying Model:
Time-invariant Model Time-varying Model
Non-linear Model

Linear Model
0
g
= u + u
l

0 sin
g
= u + u
l

Non-linear Model

Linear Model
0 sin ) t cos y
g
1 (
g

t cos y ) t ( y a , m I
sin mg sin ma I
o o
2
o o
2
o A
2
A
A A
= u e
e
+ u
e e = = =
u = u + u
l


l
l l

0 ) t cos y
g
1 (
g
o o
2
= u e
e
+ u
l

t cos y ) t ( y
0 o
e =
A
m
mg
u
F
r
F
u
l
3.4 Mathematical Representation (1)
Dynamic Representation
Time Change
Discrete Continuous Discrete Continuous
Difference
Equation
Differential
Equation
Finite
state
machine
Discrete
event
model
3.4 Mathematical Representation (2)
Nonlinearity and Linearization:
Nonlinearity in mechanical systems
x

F
Dry
Friction
x
F
Hard
Soft
Nonlinear
Spring
x
F Backlash
Hysteresis o
c
3.4 Mathematical Representation (3)
Linearization:
Linear properties:
Superposition Output response of a system to the sum of inputs
is the sum of the responses to the individual inputs.
Input Output
If r
1
(t) c
1
(t)
r
2
(t) c
2
(t)
r
1
(t) + r
2
(t) c
1
(t) + c
2
(t) Additive Property
Homogeneity The response of a system to a multiplication of the
input by a scalar.
Input Output
If r
1
(t) c
1
(t)
r
1
(t) c
1
(t) Scaling Property
3.4 Mathematical Representation (4)
Mathematical Method:
Scalar function
Taylor series expansion






Vector function

x m ) x ( f or
) x x (
dx
df
) x ( f ) x ( f e. . i
) x ( 0 ) x x (
dx
df
) x ( f ) x ( f
o
o
o
x x
o
x x
o
2
o
x x
o
o = o
=
A + + =
=
=
=
2
0 , n n
x x
n
0 , 2 2
x x
2
0 , 1 1
x x
1
n,0 2,0 0 , 1 n 2 1
) x ( 0 ) x x (
x
f
...... ) x x (
x
f

) x x (
x
f
) x ...... , x , x ( f ) x ...... , x , x ( f y
0 , n n 0 , 2 2
0 , 1 1
A +
c
c
+
c
c
+

c
c
+ = =
= =
=
3.4 Mathematical Representation (5)
1
x
f
) x , ( f
0 , 2
|
) x , x ( f f
0 , 2 0 , 1 0
=
) x , ( f
0 , 2
o
0 , 1
x o
|
=
2
x
q =
2
x
0 , 2 2
x x =
1
x
f
) , x ( f
0 , 1

) x , x ( f
0 , 2 0 , 1
) , x ( f
0 , 1
q
0 , 1
x
=
2
x
q =
2
x
0 , 2 2
x x =
For f=f(x
1
,x
2
)
| o
| o
=
c
c
) x , ( f ) x , ( f
x
f
0 , 2 0 , 2
0
1
q
q
=
c
c
) , x ( f ) , x ( f
x
f
0 , 1 0 , 1
0
2
) x x (
x
f
) x x (
x
f
) x , x ( f f
0 , 2 2
0
2
0 , 1 1
0
1
0 , 2 0 , 1

c
c
+
c
c
+ =
Experimental Method:
3.4 Mathematical Representation (6)
Dynamic Linear Model:
System
Dynamics
Linear
Model
Dynamic
Information
Prediction
Error
3.4 Mathematical Representation (7)
Classical I/O Model:
Differential form
Ex: Vibration plant





Integral form





Key points: Linearity, Causality, Relaxation

x(t) y(t) - Output
f(t) - Input : I/O
(0) x x(0), I.C.
f(t) Kx x C x M : System
=
= + +

x(t) y(t) - Output
f(t) - Input : I/O
) - g(t : System
d ) ( f ) t ( g ) t ( x
t
0
=
t
t t t =
}
y=x f(t)
Input
System
Output
C
K
M
y=x
f(t)
frictionless
3.4 Mathematical Representation (8)
Modern State-space Model:


Define system states


State equation








) t ( f
M
1
x
M
K
x
M
C
x : Ex = + +

=
=
2 1
1
x x
x x

) t ( f
M
1
0
x
x
M
C
M
K
1 0
x
x
2
1
2
1
(
(

+
(

(
(

=
(

Bf x A x + =

(0) x

I/O : Input
| | 0 1 C , x C y = =

i.e.
System:
I.C. :
Output
f(t)
) x ( x
1
) x ( x
2

state
space
x
3.4 Mathematical Representation (9)
In general
State equation :
Output equation :
I.C. :
: System states
: Control input






K: To be designed











x K ) x ( u

=
u D x C y
u B x A x

+ =
+ =
u

u D x C y
u B x A x

+ =
+ =
(0) x

u
x

3.4 Mathematical Representation (10)

State-space and I/O Model:
State-space system model







I/O system model













u b x a x a x
u b x a x a x
2 2 22 1 21 2
1 2 12 1 11 1
+ + =
+ + =

u ) b a b a ( u b x ) a a a a ( x ) a a ( x
variable x
u ) b a b a ( u b x ) a a a a ( x ) a a ( x
variable x
2 11 1 21 2 2 21 12 22 11 2 22 11 2
2
1 22 2 12 1 1 21 12 22 11 1 22 11 1
1
+ = + +

+ = + +

) dimension one ( u u ,
b
b
B ,
a a
a a
A ,
x
x
x . e . i
u
b
b
x
x
a a
a a
x
x
or
2
1
22 21
12 11
2
1
2
1
2
1
22 21
12 11
2
1
=
(

=
(

=
(

=
(

+
(

=
(

Output x
1
and / or x
2

3.4 Mathematical Representation (11)
Impulse Response:
Unit impulse function




Impulse at t=a




Shifting property
0 , 1 dt ) t (
0 t , 0 ) t (

0
> c
}
= o
= = o
c
) a t ( o
f(b)
c b a , dt ) t ( f ) b t (
c
a
=
< < o
}
c
c
1
t
t a
Area=1
3.4 Mathematical Representation (12)
Impulse Response for LTI System:
g(t) x(t) : response
) ( ) r( : Input
d ) ( r ) t ( g ) t ( x
t
0
=
t o = t
t t t =
}
Impulse is considered to generate the initial condition of a system in such a
short time that it has no other effect on the system.
Ex: Unit impulse response of 1
st
-order system
) t ( ) t ( r
) t ( r ax x
o =
= +

c
t
) t ( ) t ( r o =
at
e ) t ( x

=
a
1
0 c
1
c
1
1 x(0) C. . I
0 ax x
=
= +

t
at
e ) t ( x

=
1
3.4 Mathematical Representation (13)
r(t)
t
t
t d
1
t t
Arbitrary Input Response:
r(t) is composed of many impulses

Ex: Step response of 1
st
order system
r(t)
t
x(t)
) t ( r ax x = +