Absolute Stability

Automatic control
2. Analysis
Lesson 6
(absolute) Stability
Automatic Cintrol by Meiling CHEN
1
Stability
• Internal behavior
– The effect of all characteristic roots.
• External behavior
– The effect by cancellation of some transfer
function poles.
Automatic Cintrol by Meiling CHEN
2
Definition :
A system is internal (asymptotic) stable, if the zero-input
response decays to zero, as time approaches infinity, for
all possible initial conditions.
Asymptotic stable =>All the characteristic polynomial
roots are located in the LHP (left-half-plan)
Automatic Cintrol by Meiling CHEN
3
Definition :
A system is external (bounded-input, bounded-output)
stable, if the zero-state response is bounded, as time
approaches infinity, for all bounded inputs..
bounded-input, bounded-output stable =>All the poles
of transfer function are located in the LHP (left-halfplan)
Asymptotic stable => BIBO stable
BIBO stable=> Asymptotic stable
Automatic Cintrol by Meiling CHEN
4
System response
(i) First order system response
(ii) Second order system response
(iii) High order system response
Automatic Cintrol by Meiling CHEN
5
First order
dy
 a0 y  b0 r
dt
b0
1
Y (s) 
R( s) 
y ( 0)
s  a0
s  a0
let
r (t )  Au (t )
b0
b0
A
A
a0
a0
1
Y (s) 


y (0)
s
s  a0 s  a0
Ab0
Ab0  a0t
y (t ) 
u (t ) 
e  y ( 0) e  a 0 t
a0
a0
Automatic Cintrol by Meiling CHEN
6
Second order
d2y
dy
dr
 a1
 a0 y  b1  b0 r
2
dt
dt
dt
b1s  b0
sy(0)  ( a1  1) y (0)
Y (s)  2
R(s) 
s  a1s  a0
s 2  a1s  a0
Three cases :
(a) Two characteristic roots are real and distinct.
(b) Two characteristic roots are equal.
(c) Two characteristic roots are complex numbers.
Automatic Cintrol by Meiling CHEN
7
Two characteristic roots are real and distinct.
let
y(0)  y (0)  0
r (t )  u (t )
k3
k1
k2
Y ( s) 


s  s1 s  s2 s
y (t )  (k1e s1t  k 2 e s2t  k3 )u (t )
Automatic Cintrol by Meiling CHEN
8
Two characteristic roots are equal
let
y(0)  y (0)  0
r (t )  u (t )
k3
k1
k2
Y ( s) 


2
( s  s1 )
s  s1 s
y (t )  (k1e s1t  k 2tes1t  k3 )u (t )
Automatic Cintrol by Meiling CHEN
9
Two characteristic roots are complex numbers
let
y(0)  y (0)  0
r (t )  u (t )
k1
Y ( s) 
R( s)
2
2
(s   )  
 n2
Y ( s)  2
R( s)
2
s  2n s   n
  n
  n 1   2
Undamped natural frequency
Damping ratio
y (t )  1 
e  nt
1 2
sin(  n 1   2 t  cos 1  )
Automatic Cintrol by Meiling CHEN
10
Automatic Cintrol by Meiling CHEN
11
Automatic Cintrol by Meiling CHEN
12
Automatic Cintrol by Meiling CHEN
13
Higher-order system
 8s 2  5
Y ( s)  4
s  9s 3  37 s 2  81s  52
k3 s  k 4
k1
k2
Y ( s) 

 2
s  1 s  4 s  4s  13
Dominant root
nondominant root
Automatic Cintrol by Meiling CHEN
14
Automatic Cintrol by Meiling CHEN
15
Automatic Cintrol by Meiling CHEN
16
Automatic Cintrol by Meiling CHEN
17
Stability testing
Properties of the polynomial coefficients :
• Differing algebraic signs
7 s  5s  3s  2s  s  10
6
4
3
2
At least one RHP root
• Zero-valued coefficients
s 6  3s 5  2s 4  8s 2  3s  17
Has imaginary axis roots or RHP roots or both
• All of the same algebraic sign, non zero
8s 5  6s 4  3s 3  2s 2  7 s  10
No direct information
Automatic Cintrol by Meiling CHEN
18
Routh-Hurwitz testing
P( s )  an s n  an 1s n1    a1s  a0
an  2 an 1  an an 3
1 
an 1
sn
an
an  2
an  4  b
sn 1
an 1
an 3
an 5 
sn  2
b1
b2
sn 3
c1
c2
b3 
c3 

s
0
an  4 an 1  an an 5
b2 
an 1
an 3b1  an 1b2
c1 
b1
The number of RHP roots of P(s) is the number of
algebra sign changes in the elements of the left column of the array.
Automatic Cintrol by Meiling CHEN
19
Example 1
P ( s )  2 s 4  3s 3  5 s 2  2 s  6
s
4
2
5
s
3
s
2
2
18  0
6
3
s
1
3
15  4 11

3
3
 32
11
6
s0
6
Two roots in the RHP
Automatic Cintrol by Meiling CHEN
20
Example 2
P ( s )  s  2 s  3s  4 s  1
4
3
4
1
3 1
s3
2
4
s2
1
1
s1
2
s0
1
s
2
no root in the RHP
Automatic Cintrol by Meiling CHEN
21
Example 3
P ( s )  3s  6 s  2 s  4 s  5
4
3
2
4
3
2
s3
6
4
s2
A
0
5
B ( 1) 5
s
s1
s0
5
0
5
n
s
2
5 5
s
1
10
s
0
5
n 移位次數移
至0消失為止
Two roots in the RHP
Automatic Cintrol by Meiling CHEN
22
Example 4
P ( s )  s 5  2 s 4  8s 3  11s 2  16 s  12
s5
1
8
s4
2
11 12
s
3
s
2
s
1
s
0
16
factor
2.5 10
3
0
12
3s  12
2
0
Automatic Cintrol by Meiling CHEN
23
P ( s )  s 5  2 s 4  8s 3  11s 2  16 s  12
s5
1
8
s4
2
11 12
s
3
s
2
s
1
6
s
0
12
16
2.5 10
3
12
d
2
(3s  12)  6 s
ds
no root in the RHP
Automatic Cintrol by Meiling CHEN
24
Example 5
R(s) +
-
k
s
10
s 2  2  10
Y (s )
1
s2
Y ( s)
10k

R( s) s( s 2  s  10)( s  2)  10k
Automatic Cintrol by Meiling CHEN
25