Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Modern Control Systems

Dr. Mohammad Al Janaideh


Eigenvalues
• Eigenvalues of an n×n matrix
• The eigenvalues of an n×n matrix A are the roots of the
characteristics equation:
I  A  0
0 1 0   1 0 
   
A 0 0 1 , I  A   0   1 ,
 6  11  6  6 11   6
   

 3  62  11  6
1  1, 2  2, 3  3
 (  1)(  2)(  3)  0
Diagonalization
aMatrix diagonalization is the process X  AX  Bu
of taking a square matrix and
converting it into a special type of y  CX
matrix-also-called diagonal matrix-
that shares the same fundamental Pz  APz  Bu
properties of the underlying matrix.
y  CPz
Diagonalizing a matrix is also
equivalent to finding the matrix's
X  APz  Bu
eigenvalues, which turn out to be y  CPz
precisely the entries of the
diagonalized matrix. z  P 1 APz  P 1Bu
y  CPz
Diagonalization
0 1 0 ... 0 
1  X  Pz
0 0 1 ... 0 
 
A . . . ... . 
A has n distinct eigenvalues
 
 0 0 0 ... 1 
 a n  an 1  an  2 ...  a1 
 

 1 1 1 ... 1  1 0 0 ... 0 


 n  0 
 1
2 3 ...
 0 ... 0 
2 2 2 2
 2 

P 1  2
 3
... n  P 1 AP   0 0 3 ... 0 
   
 . . . ... .  . . . ... . 
n 1 n 1 n 1 ... nn1  0 0 0 ... n 
 1 2 3   
Example (1)
 x   0 1 0   x1  0 x 
A  x    0 0 1   x   0u
    2     
 1
y  1 0 0  x2 
 x   6  11  6  x  6 x 
     3     3 

(1) The Eigenvalues matrix A are : λ1= -1 , λ2 = -2, λ3= -3


(2) The matrix P are : λ1= -1 , λ2 = -2, λ3= -3
1 1 1 1 1 1
   
P   1 2 3  P   1  2  3
2 22 32  1 4 9 
 1   
Example (1)
X  PZ Z  P 1 APZ  P 1 Bu
1
z 
y  CPZ
 1
Z   z2 
z   3 2.5 0.5
 3   
P 1   3  4  1
 1 1.5 0.5
X  AX  Bu  
y  CX
0 
 
PZ  APZ  Bu
B  0 
6 

C1 0 0 
 
y  CPZ
Example (1)
1
 1 0 0 3
1    
P AP   0  2 0  1
P B    6
0 0  3 3
   

λ1= -1 , λ2 = -2, λ3= -3 


CP  1 1 1 
The Eigenvalues P 1 AP identical to the eigenvalues of the A.
(Proof)
The Eigenvalues P 1 AP identical to the eigenvalues
of the A.

Proof:
Example (2)
 x   0 1 0   x1  0 x 
A  x    0 0 1   x   0u
    2     
 1
y  1 0 0  x2 
 x   6  11  6  x  6 x 
     3     3 

G(s)  C (sI  A) 1 B
1
  s 0 0  0 1 0   0 
   
 

 
G ( s )  1 0 0  0 s 0    0 0 1   0 
  
  
 0 0 s   6  11  6  6
     
6
G ( s)  3 2
s  6s  11s  6
6 A B C
G ( s)  3   
s  6s 2  11s  6 s 1 s  2 s  3

3 6 3
G ( s)   
s 1 s  2 s  3

 1 0 0  3

A 0 2 0 
0


 
B    6 
C1 1 1 
0  3  3
   
Transformation of System Models with
Matlab
Using Matlab we can transform the system model from
transfer function to state space.
Y (s) num

U (s) den

We use the following Matlab command to transform the


transfer function to state space :
[A,B,C,D]=tf2ss(num,den)
We use the following Matlab command to transfer the
state space to transfer function:
[num,den]=ss2tf(A,B,C,D)
Example (1)
a 2s 2  s  9 A=
G ( s)  -2 -8 -8 -2
s 4  2 s 3  8 s 2  8s  2 1 0 0 0
0 1 0 0
0 0 1 0
B=
>> num=[ 2 1 9]; 1
>> den=[ 1 2 8 8 2]; 0
>>[A,B,C,D]=tf2ss(num,den) 0
0
C=
0 2 1 9

D=
0
Example (2)
2 8 8 2
a 1 0 0 0 
1  C  0 2 1 9
0 
 
A  B  D0
0 1 0 0 0 
   
 0 0 1 0  0

>>A=[-2 -8 -8 -2;1 0 0 0; 0 1 0 0; 0 0 1 0];


>>B=[1;0;0;0];
>>C=[0 2 1 9] ; 2
2 s s9
>>D=0; G ( s)  4 3 2
>>[num,den]=ss2tf(A,B,C,D) s  2 s  8 s  8s  2
num =
0 -0.0000 2.0000 1.0000 9.0000
den =
1.0000 2.0000 8.0000 8.0000 2.0000

You might also like