1.

System Description and Analysis

Mathematical Modeling review

Generalized Energy Variables

Translational Rotational
Domain >> Electrical Hydraulic Thermal
mechanical mechanical
e effort F force, τ torque, V voltage, P pressure, T
variable newtons newton-m volts N/m2 temperature
v velocity, ω angular i current, F flow, Q heat flow,
f flow variable m/sec velocity, 1/sec amperes m3/sec watts
p integral of H angular
effort, generic P momentum no name no name no name
momentum
momentum
x integral of θ angular q charge, E energy,
flow, generic X displacement V volume
displacement coulombs joules
displacement
rotational
R resistive damper
damper
resistor constriction insulator

torsional balloon- thermal

C capacitive spring
spring
capacitor
thing reservoir
there is
I inertia mass flywheel inductor
none
TF lever gear pair transformer
transformer

Analogous Quantities

ElectricalQuantity Mechanical (Force Voltage)

Voltage, e Force, f
Current, i Velocity, v
Resistance, R Friction, B
Compliance, 1/K
Capacitance, C
(Inverse spring constant)
Inductance, L Mass, M
Transformer, N1:N2 Lever, L1:L2

1/21
 1. System Description and Analysis

Analogous Equations
Mechanical
Electrical
Analog II
Equation
(Force Voltage)

2/21
 1. System Description and Analysis

Electrical Systems

Inductive: Inductor Capacitive: Capacitor Resistive: Resistor

i(t) L C i(t) R
i(t)
+ v(t) _ _ + v(t) _
+
v(t)
di ( t ) 1 t v ( t ) = Ri ( t )
v (t ) = L V ( t ) = ∫ i ( t )dt
dt C −∞ 1
1 t dv ( t ) i (t ) = v (t )
i (t ) = v ( t ) dt i (t ) = C R
L ∫−∞ dt
Stores Kinetic Energy Stores Potential Energy Dissipates Power
1 1 P = v (t ) i (t )
K .E. = Li ( t ) P.E. = Cq ( t )
2 2

2 2

1.1.1) Mechanical Systems (Translational)

Inductive: Mass Capacitive: Spring Resistive: Damper
v(t) v(t) v(t)
F(t) F(t) F(t)
m

dv ( t ) t
F ( t ) = k ∫ v ( t )dt F ( t ) = bv ( t )
F (t ) = m −∞
dt 1
1 1 dF ( t ) v (t ) = F (t )
t
v ( t ) = ∫ F ( t ) dt v (t ) = b
m −∞ k dt
Stores Kinetic Energy Stores Potential Energy Dissipates Power
1
K .E. = mv ( t )
2 1
P.E. = kx ( t )
2 P = F (t ) v (t )
2 2

1.1.2) Mechanical Systems (Rotational)

Inductive: Moment of Capacitive: Spring Resistive: Damper
Inertia
T(t) T(t) T(t)
I
ω(t) ω(t) ω(t)
dω ( t ) t
T ( t ) = k ∫ ω ( t )dt T ( t ) = bω ( t )
T (t ) = I −∞
dt 1
1 1 dT ( t ) ω (t ) = T (t )
ω (t ) =
t
T ( t ) dt ω (t ) = b
I ∫−∞ k dt
Stores Kinetic Energy Stores Potential Energy Dissipates Power

3/21
 1. System Description and Analysis
K .E. =
1
Iω (t )
2 1
P.E. = kθ ( t )
2 P = T (t )ω (t )
2 2

State Space and Transfer Function


x ( t ) Ax ( t ) + Bu ( t )  State Equation
=
y ( t ) Cx ( t ) + Du ( t )  Output Equation
=

Where
 x1 ( t )   u1 ( t )   y1 ( t ) 
     
 x2 ( t )   u2 ( t )   y2 ( t ) 
x (t ) =  n states u (t ) = k inputs y (t ) = m outputs
        
     
 xn ( t )  uk ( t )   ym ( t ) 

A  State Matrix [n×n] [Note: poles of the system  eigenvalues of A]

B  Input Matrix [n×k]
C  Output Matrix [m×n]
D  Direct transmission Matrix [m×k]

Taking Laplace transform of state space equations

sX ( s ) − x ( 0 )= AX ( s ) + BU ( s )
Y ( s ) CX ( s ) + DU ( s )
=

( sI − A) X ( s ) = x ( 0 ) + BU ( s )
X ( s ) = ( sI − A ) x ( 0 ) + ( sI − A ) BU ( s )
−1 −1

  
Zero Input Response Zero State Response

Y ( s=) C ( sI − A) x ( 0 ) + ( sI − A) BU ( s ) + DU ( s )

−1 −1

Y ( s ) =C ( sI − A ) x ( 0 ) + C ( sI − A ) B + D  U ( s )
−1 −1

   
Zero Input Response Zero State Response

Zero Input Response: In electrical circuit theory, the Zero Input Response or ZIR is the
behavior or response of a circuit with zero inputs. The ZIR results only from the initial
state of the circuit and not from any external drive. The ZIR is also called the natural
response, and the resonant frequencies of the ZIR are called the natural frequencies.

Zero State Response: In electrical circuit theory, the Zero State Response or ZSR is the
behavior or response of a circuit with initial state of zero. The ZSR results only from the

4/21
 1. System Description and Analysis
external inputs or driving functions of the circuit and not from the initial state. The ZSR
is also called the forced or driven response of the circuit.

1.2) Transfer function from State Space Model

Transfer function is defined as
H ( s ) = Y ( s )U ( s )
−1

initial conditions = 0

From,
Y ( s ) =C ( sI − A ) x ( 0 ) + C ( sI − A ) B + D  U ( s )
−1 −1

   
Zero Input Response Zero State Response

H ( s ) = Y ( s )U ( s )
−1

initial conditions = 0

=C ( sI − A ) B + D
−1

1.3) Single Input Single Output (SISO) system

H ( s )  [1×1] matrix i.e. scalar

1.3.1) Example: Mass Damper

Input u = F and Output y = v
i:v L:m
b
F
m V:F R:b
v
Mathematical Model:

F m v + bv
=
State: x1 = v
 1
x1= v= ( F − bv ) y1 = v
m
b 1 y1 = x1
− x1 + u1
=
m m

State space representation:


 b 1
x= − m  x +  m  u =y [1] x + [ 0] u
   

Transfer function:
H ( s ) =C ( sI − A ) B + D
−1

5/21
 1. System Description and Analysis
−1
 b  1
H
= ( s ) [1]  s [1] −  −     + [ D ]
  m m
1
= m
b
s+
m

1
H (s) =
ms + b

1.3.2) Example: Mass Spring Damper

Input u = F and Output y = x
i:v L:m C:1/k
b
F
m R:b
V:F
x k
Mathematical Model:

di 1
dt c ∫
V =L + idt + Ri
o
dq 1 o o
V =L + ∫ q dt + R q
dt c
oo 1 o
V = L q+ q + R q
c
Mechanical form of the equation
 
F = m x + kx + b x

State: x1 = x , x2 = x
  1 
 y1 = x
x1 =x =  F − b x − kx 
m  y1 = x2
b k 1
=− x1 − x2 + u1
m m m
 
x 2= x= x1

State space representation:

 b k 1

 − − 
=x m m x + mu =y [0 1] x + [0] u
   
 1 0  0

6/21
 1. System Description and Analysis
Transfer function:
H ( s ) =C ( sI − A ) B + D
−1

−1 −1
  b k   1   b k  1
 1 0   − −      s+ 
=H ( s ) [ 0 1] s  −
 0 1   m m = m + [ 0] [ 0 1]  m m  m  + [ 0]
       
  1 0    0   −1
 s   0
 k 
 s − 1  1
1 m   1  b  
= [ 0 1] =  m 1 s+  m
 b k  b    s 2 + b s + k  m  
s s +  + 1 s+  0 0
 m  m  m  m m
1
= m
b k
s2 + s +
m m
1
H (s) = 2
ms + bs + k

1.3.3) Flexible connection [4th order system]

Input u = F and Output y = y1

b
F
m2 m1
k
y2 y1

Mathematical Model:

   
m2 y2 =F − b  y2 − y1  − k ( y2 − y1 )
 

  

m1 y1 =−b  y1 − y2  − k ( y1 − y2 )
 
 
States: x1 = y1 x2 = y1 x3 = y2 x4 = y2
 
x=
1 y=
1 x2
  1      
x2 =y1 =  −b  y1 − y2  − k ( y1 − y2 ) 
m1    
k b k b
− x1 −
= x2 + x3 + x4
m1 m1 m1 m1
 
x=
3 y=
2 x4

7/21
 1. System Description and Analysis
  1   



x4 = y2 = u − b  y2 − y1  − k ( y2 − y1 ) 
m2    
k b k b 1
= x1 + x2 − x3 − x4 + u
m2 m2 m2 m2 m2

State space representation:

 0 1 0 0 
 k   0 
− b k b   0 
−
  1 m m m m1 
 
x 
1 1
x+ 0 = u y [1 0 0 0] x + [ 0] u
 0 0 0 1   
 k b k b   1 
 − −   m2 
 2m m 2 m 2 m 2 

M-file for calculating Transfer system in Matlab (Symbolic Manipulation)

H ( s ) =C ( sI − A ) B + D
−1

msdm.m
%An example of solution of a mass spring mass system
%(a 4th order system)using symbolic toolbox of Matlab
clear all;
syms m1 m2 b k; %defining parameters
syms A B C D; %System Matrices
syms TF s; %Transfer function and laplace variable s
%SYMS(MATLAB command for creating symbolic
object

%Defining state state equations obtained from analysis

A=[0 1 0 0;-k/m1 -b/m1 k/m1 b/m1;0 0 0 1;k/m2 b/m2 -k/m2 -b/m2];
B=[0; 0; 0; 1/m2];
C=[1 0 0 0];
D=[0];

disp ('state state representation of System')

A, B, C, D

%Transfer function given by C*inv[sI-A]*B+D

TF= C*inv((s*eye(length(A))-A))*B+D;

disp ('Transfer Function of the system [TF]')

TF=simple (TF); %simplifying the transfer function
pretty (TF); %pretty prints the symbolic expression
%Verification checking of poles of the system
P1=solve (1/TF,s); %poles obtained from transfer function
P2=eig (A); %poles obtained from state state

disp ('poles obtained from Transfer Function [P1]')

P1=simple (P1); %Simplifying
pretty (P1);

disp ('poles obtained from State Space model [P2]')

P2=simple (P2); %Simplifying
pretty (P2);

8/21
 1. System Description and Analysis
Matlab command screen
>> msdm
state state representation of System
A =
[ 0, 1, 0, 0]
[ -k/m1, -b/m1, k/m1, b/m1]
[ 0, 0, 0, 1]
[ k/m2, b/m2, -k/m2, -b/m2]
B =
[ 0]
[ 0]
[ 0]
[ 1/m2]
C =
1 0 0 0
D =
0
Transfer Function of the system [TF]
k + s b
---------------------------------------------
2 2
s (s m1 m2 + (m1 b + b m2) s + k m2 + m1 k)
poles obtained from Transfer Function [P1]
[ 0 ]
[ ]
[ 0 ]
[ ]
[ 2 2 1/2]
[ m1 b + b m2 - ((m1 + m2) (m1 b - 4 m1 k m2 + b m2)) ]
[- 1/2 --------------------------------------------------------]
[ m1 m2 ]
[ ]
[ 2 2 1/2]
[ m1 b + b m2 + ((m1 + m2) (m1 b - 4 m1 k m2 + b m2)) ]
[- 1/2 --------------------------------------------------------]
[ m1 m2 ]
poles obtained from State Space model [P2]
[ 0 ]
[ ]
[ 0 ]
[ ]
[ 2 2 1/2]
[ m1 b + b m2 - ((m1 + m2) (m1 b - 4 m1 k m2 + b m2)) ]
[- 1/2 --------------------------------------------------------]
[ m1 m2 ]
[ ]
[ 2 2 1/2]
[ m1 b + b m2 + ((m1 + m2) (m1 b - 4 m1 k m2 + b m2)) ]
[- 1/2 --------------------------------------------------------]
[ m1 m2 ]

Transfer function of the system is given by

Y (s) bs + k
H (s) = 1
F ( s ) s  m1m2 s + ( m1 + m2 ) bs + ( m1 + m2 ) k 
2 2

And four poles of the system are given by

1 m1b + m2b ± − ( m1 + m2 ) ( −m1b + 4m1m2 k − m2b )
2 2

0, 0 and −
2 m1m2

9/21
 1. System Description and Analysis
1.4) Multiple Input Multiple Output (MIMO) system
Y ( s ) = H ( s )U ( s )
H ( s )  [m×k] matrix
 H11 ( s ) H12 ( s )  H1k ( s ) 
 
 H 21 ( s )   
H (s) =
    
 
 H m1 ( s )   H mk ( s ) 
Where,
Yi ( s )
H ij ( s ) =
U j (s)
U k ( s ) =0∀k ≠ j

1.4.1) Example: Flexible connection [4th order system] multiple output

Input u = F and Output y = [ y1 y2 ]
T

b
F
m2 m1
k
y2 y1
States (same as previous):
 
x1 = y1 x2 = y1 x3 = y2 x4 = y2
State space representation:
 0 1 0 0 
 k  0 
− b k b   0 
−
  m1 m1 m1 m1    1 0 0 0  0 
x   x +  0= u y   x +  u
 0 0 0 1   
1 
0 0 1 0  0 
 k b k b  
 − −   m2 
 m2 m2 m2 m2 

H ( s ) =C ( sI − A ) B + D
−1

 bs + k 
 s 2  m m s 2 + m + m bs + m + m k  
  1 2 ( 1 2) ( 1 2) 
H (s) =  
m1s 2 + bs + k
 2 
 s  m1m2 s 2 + ( m1 + m2 ) bs + ( m1 + m2 ) k  

Y1 ( s ) bs + k
H11
= (s) = 2
F ( s ) s  m1m2 s + ( m1 + m2 ) bs + ( m1 + m2 ) k 
2

10/21
 1. System Description and Analysis
Y2 ( s ) m1s 2 + bs + k
H12
= ( )
s =
F ( s ) s 2  m1m2 s 2 + ( m1 + m2 ) bs + ( m1 + m2 ) k 

bs + k
F(s) Y1(s)
s  m1m2 s + ( m1 + m2 ) bs + ( m1 + m2 ) k 
2 2

m1s 2 + bs + k Y2(s)
s 2  m1m2 s 2 + ( m1 + m2 ) bs + ( m1 + m2 ) k 

1.4.2) Example: Armature Controlled DC Motor

Input u = [ e TL ] , Output y = [ω θ ]
T T

Ra = armature resistance
La = armature inductance
ia = armature current
if = field current [Fixed field]
e = applied armature voltage
TM = motor torque
θ = angular displacement of shaft
ω = angular speed
J = moment of inertia
fo = equivalent viscous friction coefficient of motor

air gap flux, φ = K f if K f field constant

torque TM = φK 1ia = K T ia KT  armature constant
back emf eb = K b ω K b  motor constant

The balance equations give (differential equations):

dθ (t ) di (t )
1) eb (t ) = K bω (t ) = K b 2) e(t ) = eb (t ) + La a + Ra ia (t )
dt dt
d θ (t )
2
4) TM ( t ) = KT ia ( t )
3) J =T M ( t ) − f oω ( t ) − TL ( t )
dt 2

11/21
 1. System Description and Analysis

State Space Representation

u = [ e TL ] , y = [ω θ ] and x = [ia ω θ]
T T T

 Ra K 1  Ra Kb  1 
x1 (t ) = − x1 (t ) − b x 2 (t ) + u1 (t ) − L − 0 L 0 
La La La La
 a   a 
 K f 1   K f   1
x 2 (t ) = T x1 (t ) − o x 2 (t ) − u 2 (t ) x=  T − o 0 x +  0 − u
J J J

 J J  
J

x 3 (t ) = x 2 (t )  0 1 0 0 0 
   
   
y1 ( t ) = x2 ( t ) 0 1 0 0 0 
y=  x+ u
y2 ( t ) = x3 ( t ) 0 0 1 0 0 

M-file for calculating Transfer system in Matlab (Symbolic Manipulation)

H ( s ) =C ( sI − A ) B + D
−1

dc_motor.m
%An example of solution of Armature Controlled DC Motor system
%(a MIMO system)using symbolic toolbox of Matlab

clear all;
syms Ra La Kb KT fo J; %defining parameters
syms A B C D; %System Matrices
syms TF s; %Transfer function and laplace variable s

%Defining state state equations obtained from analysis

A=[-Ra/La -Kb/La 0; KT/J -fo/J 0; 0 1 0];
B=[1/La 0; 0 -1/J; 0 0];
C=[0 1 0;0 0 1];
D=[0 0;0 0];

disp ('state state representation of System')

A, B, C, D

%Transfer function given by C*inv[sI-A]*B+D

TF= C*inv((s*eye(3)-A))*B+D;

disp ('Transfer Function of the system [TF]')

TF=simple (TF); %simplifying the transfer function
pretty (TF);

Matlab command screen

>> dc_motor
state state representation of System
A =
[ -Ra/La, -Kb/La, 0]
[ KT/J, -fo/J, 0]
[ 0, 1, 0]
B =
[ 1/La, 0]
[ 0, -1/J]
[ 0, 0]
C =

12/21
 1. System Description and Analysis
0 1 0
0 0 1
D =
0 0
0 0
Transfer Function of the system [TF]
[ KT s La + Ra]
[---- - ---------]
[ %1 %1 ]
[ ]
[ KT s La + Ra]
[---- - ---------]
[s %1 s %1 ]
2
%1 := s La J + (La fo + Ra J) s + KT Kb + Ra fo

 KT − ( sLa + Ra ) 
 2 
 s La J + ( La f o + Ra J ) s + KT K b + Ra f o s La J + ( La f o + Ra J ) s + KT K b + Ra f o 
2

H (s) =  
KT − ( sLa + Ra )
 
(
 s s La J + ( La f o + Ra J ) s + KT K b + Ra f o

2
) ( )
s s 2 La J + ( La f o + Ra J ) s + KT K b + Ra f o 

ω (s) KT
H11 ( s )
= = 2
E ( s ) T s =0 s La J + ( La f o + Ra J ) s + KT K b + Ra f o
L ( )
ω (s) − ( sLa + Ra )
H12 ( s )
= = 2
TL ( s ) E s =0 s La J + ( La f o + Ra J ) s + KT K b + Ra f o
( )

θ (s) KT
H 21 ( s )
= =
L( )
2
(
E ( s ) T s =0 s s La J + ( La f o + Ra J ) s + KT K b + Ra f o )
θ (s) − ( sLa + Ra )
H 22 ( s )
= =
(
TL ( s ) E s =0 s s La J + ( La f o + Ra J ) s + KT K b + Ra f o
( )
2
)

13/21
 1. System Description and Analysis
KT
E(s) ω(s)
s La J + ( La f o + Ra J ) s + KT K b + Ra f o
2

− ( sLa + Ra )
s 2 La J + ( La f o + Ra J ) s + KT K b + Ra f o

KT
(
s s La J + ( La f o + Ra J ) s + KT K b + Ra f o
2
)

− ( sLa + Ra )
TL(s) θ(s)
(
s s La J + ( La f o + Ra J ) s + KT K b + Ra f o
2
)

KT ω(s) 1
E(s) θ(s)
s La J + ( La f o + Ra J ) s + KT K b + Ra f o
2
s

− ( sLa + Ra )
TL(s)
s 2 La J + ( La f o + Ra J ) s + KT K b + Ra f o

14/21
 1. System Description and Analysis

Hydraulic System
Qin

Qin
Qout h
h Qout Tank

Area A

dM
= ∑ Mass Flow
dt
dM
= Qin − Qout
dt
d ρV
= Qin − Qout
dt
dh
ρ A = Qin − Qout
dt 
h
Qin 1 h
dh 1 ∫
= ( Qin − Qout ) ρA
dt ρ A Qout

Laminar Flow Turbulent Flow

Qout = Kh Qout = K h
dh 1 dh 1
=
dt ρ A
( Qin − Kh ) =
dt ρ A
(
Qin − K h )
K  [m2/s] K  [m2.5/s]

15/21
 1. System Description and Analysis
Linearization

1.4.3) Linearization (Single variable equation)

Linear equation  superposition theorem applies
y = 2x

Nonlinear equation
y = x2

Laplace transform applies only for linear equations

Taylor Series of y = f ( x )
df 1 d2 f 1 dn f
y lim f ( x0 ) + ( 0) ( x − x0 ) +  + ( x − x0 )
2 n
= x − x + +
x → x0 dx
= x = x0 2! dx 2 x x=
n ! dx n x x0
0

Neglecting second and higher order differential equations, gives a linear equation:
df
y = lim f ( x0 ) + ( x − x0 )
x → x0 dx x = x0
df
y=
− yw ( x − xw ) where yw , xw  working point
dx x = xw

df
∆y
= ∆x
dx x = xw

Example:
y = x 2 linearization at xw = 1 y = x 2 linearization at xw = 2
2 2
yw x=
= w 1 yw x=
= w 4
− 1 2 xw ( x − 1)
y= − 4 2 xw ( x − 2 )
y=
1 2 ( x − 1)
y −= y − 4= 4 ( x − 2 )
∆y = 2∆x ∆y = 4∆x

16/21
 1. System Description and Analysis
y = x2
2
∆y = 2∆x

1
( xw , yw ) = (1,1)

0
0 1

1.4.4) Linearization (Multiple variable equation)

y = f ( x1 , x2 )
Linear equation:
∂f ∂f
y − yw
= ( x1 − x1w ) + ( x2 − x2 w )
∂x1 x , x ∂x2 x1 w , x2 w
1w 2w

∂f ∂f
=∆y ∆x1 + ∆x2
∂x1 x1 w , x2 w
∂x2 x1 w , x2 w

Example: y1= x1 × x2 at x1w = 1 and x2 w = 2

yw = x1w × x2 w = 2
− 2 x2 w ( x1 − 1) + x1w ( x2 − 2 )
y=
2 2 ( x1 − 1) + 1( x2 − 2 )
y−=

17/21
 1. System Description and Analysis
Example: Pendulum

Moment of inertia J = ml 2
Torque
= T mg sin (θ ) × l

l J θ = −T
θ

ml 2 θ = −mgl sin (θ )
 g
θ = − sin (θ )
mg×sin(θ) l

mg mg×cos(θ)

θ = f (θ )
Linearization at θ w = 0 radians


For working point: θ = 0 (steady state)

 ∂f  g
θ −=
0 (θ − 0 ) θ − 0 =− cos (θ w )(θ − 0 )
∂θ l
 g
θ= − θ
l

1.4.5) Linearization of State Space Equation

Linear State Space:

x Ax + Bu and =
= y Cx + Du
Non linear State Space:

x = f ( x , u ) and y = g ( x , u )

Example:
 
 x1  =  x1 ( x2 + u1 )  y= x1 + u1
    x1 + x23  1

 x2 
at u1w = 2


For working point: x w = 0 (steady state)
=0 x1w ( x2 w + 2 ) x1w = 0 or 8
0 x1w + x2 w3
= =x2 w 0 or −2
y1w = x1w + u1w = 10

18/21
 1. System Description and Analysis
 
x1 − x1w = ( x2 w + u1w )( x1 − x1w ) + x1w ( x2 − x2 w ) + x1w ( u1 − u1w )

∆ x1 = 0∆x1 + 8∆x2 + 8∆u1
 
x 2 − x 2 w = 1( x1 − x1w ) + 3 x2 w 2 ( x2 − x2 w ) + 0 ( u1 − u1w )

∆ x 2 =1∆x1 + 12∆x2 + 0∆u1

∆y1 = ∆x1 + ∆u1

Linearized State Space at u1w = 2

 0 8  8 
∆x 
=  ∆x +   ∆u ∆y [1 0] ∆x + [1] ∆u
=
1 12  0 

1.4.6) Generalized Linearization of State Space Equation

Non linear state space:

x = f ( x, u )
y = g ( x, u )
n states, k inputs and m outputs
 n state functions f(t) [f1(t) … fn(t)]and m output functions g(t) [g1(t) … gm(t)]
 ∂f1 ∂f1   ∂f1 ∂f1 
 ∂x  ∂x   ∂u  ∂u 
  1 n
  1 k

=∆ x      ∆x +      ∆u
   
 ∂f n  ∂f n   ∂f n  ∂f n 
 ∂x1 ∂xn   ∂u1 ∂uk 

 ∂g1 ∂g1   ∂g1 ∂g1 

 ∂x    ∂u 
∂xn ∂uk 
 1   1 
=∆y      ∆x +      ∆u
   
 ∂g m 
∂g m   ∂g m 
∂g m 
 ∂x1 ∂xn   ∂u1 ∂uk 

Example:
   2
 x1  =  x1 sin ( u1 ) + 8 x2  → f1 − x12 + x2 → g1
y1 =
    x1 x2 + cos ( u1 ) + 1 → f 2
 x2 
π
at u1w =
4

19/21
 1. System Description and Analysis
1
=0 x1w 2 + 8 x2 w
2 x1w = 2.68

1 x2 w = −0.64
0= x1w x2 w + +1
2
− x1w 2 + x2 w =
y1w = −7.83

 ∂f1 ∂f1   ∂f1 

  ∂x ∂x2   ∂u 
=∆x  1
 ∆x +  1  ∆u
 ∂f 2 ∂f 2   ∂f 2 
 ∂x ∂x2   ∂u 
 1  1
 2 x1w sin ( u1w ) 8 
  x1w 2 cos ( u1w ) 
=∆x   ∆x +   ∆u
 x2 w x1w   sin ( u1w ) 
  3.79 8   5.09 
=∆x   ∆x +   ∆u
 −0.64 2.68  −0.707 

 ∂g ∂g1   ∂g1 
=∆y  1  ∆x +   ∆u ∆y = [ −2 x1w 1] ∆x + [ 0] ∆u
 ∂x1 ∂x2   ∂u1 
∆y = [ −5.36 1] ∆x + [ 0] ∆u

Example: Water Tank with Turbulent flow

input: [u1 ] = [Qin ] and output: [ y1 ] = [ h ]
dh 1
=
dt ρ A
(
Qin − K h )
state: [ x1 ] = [ h ]
 1 1
x1 =
−K x1 + u → f1 y= x1 → g1
ρA ρA 1 1

  ∂f   ∂f    1 1   1 
∆=x  1  ∆x +  1  ∆u ∆ x = − K  ∆x +   ∆u
 ∂x1   ∂u1   ρA 2 h   ρ A
 ∂g   ∂g 
∆y  1  ∆x +  1  ∆u
= ∆y = [1] ∆x + [ 0] ∆u
 ∂x1   ∂u1 

Example: Non-linear Spring

y non linear Input [u1 ] = [ F ]
spring
Output [ y1 ] = [ y ]
F m Non-linear spring
= k1 y + k2 y 3
Fspring
b
Non-linear model

20/21
 1. System Description and Analysis
 
F = m y + b y + k1 y + k2 y 3
T
 
States: [ x1 x2 ] =  y y 
T

 
 b  k k 1
y= − y− 1 y − 2 y3 + F
m m m m
 b k k 1 
x1 = − x1 − 1 x2 − 2 x23 + u1 → f1 x=
2 x1 → f 2
m m m m

 ∂f1 ∂f1   ∂f1 

  ∂x ∂x2   ∂u 
=∆x  1  ∆x +  1  ∆u
 ∂f 2 ∂f 2   ∂f 2 
 ∂x ∂x2   ∂u 
 1  1
 b k1 k2  1
−  − − 3 x2 w 2 
=∆x  m m m ∆x +  m  ∆u
   
 0 1  0
y=
1 x2 → g1
 ∂g ∂g1   ∂g1 
=∆y  1  ∆x +   ∆u
∂
 1x ∂x2  ∂u
 1
∆y [ 0 1] ∆x + [ 0] ∆u
=

21/21