You are on page 1of 16

Outline

Stability of Digital Control Systems

M. Sami Fadali Professor of Electrical Engineering UNR

 

Asymptotic stability.

Input-output stability.

Stability conditions.

Routh-Hurwitz criterion.

Jury test.

Nyquist criterion and stability margins.

1

2

Asymptotic Stability

Response due to any initial conditions decays to zero asymptotically in the steady state, i.e. the response due to the initial conditions satisfies

_|Å

(

y k

) 0

k 

Marginal Stability: response due to any initial conditions remains bounded but does not decay to zero.

Bounded-Input-Bounded-Output (BIBO) Stability

Response due to any bounded input remains bounded.

(

uk

)

b

u

,

k

 

b

0  

y kb

(

)

y

u

,

k

 

b

0  

y

012

 

,

,

, ...

012

,

,

, ...

Stable Z-Domain Pole Locations

Sampled exponential and its z-transform with p real or complex

p

k

,

k

0,1, 2, ...

 Z

_|Å

k 

k

p

0,

1,

p

p

, p

1

1

1

z

z

p

5

Time Sequence

f

(

k

)

n

A p

i

k

i

i

1

,

k

0,1, 2,

...

Z



F

(

z

)

n

A i
A
i

i

1

z

z p

i

  • Bounded sequence for poles in the closed unit disc (i.e. on or inside the unit circle).

  • Sequence decays exponentially for poles in the open unit disc (i.e. inside the unit circle).

  • Unbounded sequence for repeated poles on the unit circle.

  • For real time sequences poles and partial fraction coefficients are either real or complex conjugate pairs.

6

Theorem: Asymptotic Stability

In the absence of pole-zero cancellation, a LTI digital system is asymptotically stable if its transfer function poles are in the open unit disc and marginally stable if the poles are in the closed unit disc with no repeated poles on the unit circle.

7

Proof

(

y k

n

)

a

n 1

(

y k

 

n

1)

...

(

a y k

1

1)

a

0

(

y k

)

(

b u k

m

m

)

(

b u k

m

m

1)

...

(

b u k

1

1)

b u

0

(

k

)

k 0,1, 2, ...

Response of the system due to ICs y(0), y(1), , y(n)

Y

(

z

)

N

(

z

)

z

n

a

n 1

n 1

z

 

...

az

1

a

0

  • 1. Denominator of the output z-transform = denominator of the z-transfer function for no pole-zero cancellation.

  • 2. Poles of Y(z) = poles of transfer function.

  • 3. Y(z) due to initial conditions is bounded for system poles in the closed unit disc unit (no repeated poles on the unit circle) & decays exponentially for system poles

in the open unit disc (inside the unit circle).

8

Example

Solution

Determine the asymptotic stability of the following systems:

Use Theorem 1 can (a) and (b) without pole- zero cancellation.

Ignore zeros, (do not affect response due ICs)

 

H

4

z

2

b

H

 

4

z

0.2

  • a) Pole outside the unit circle unstable

 

a

)

(

z

)

z

2



z

0.1

)

(

z

)

z

0.2



z

0.1

 
  • b) All poles inside the unit circle asymptotically stable.

 

)

H

(

z

)

5

z

0.3

d

)

H

(

 

)

8

z

0.2

c

z

0.2



z

0.1

z

z

0.1

z

1

  • c) All poles inside the unit circle asymptotically

  • d) One pole on the unit circle marginally stable.

 

stable.

 

9

10

 

BIBO Stability

 

Theorem 4.2:BIBO Stability

 

k

( ) 

yk

(

hk

iui

)

( ),

 

k

012

,

,

, ...

 

i 0

Is system BIBO stable if its impulse response h(k) is bounded? NO

A discrete-time linear system is BIBO stable if and only if its impulse response sequence is absolutely summable i.e.

Counterexample: bounded & strictly positive impulse response

0

b

h

1

(

h k

)

b

h

2

 

,

k

0,1, 2, ...

y ( k )

k

(

h k

i

0

)

i u

( )

i

b

h 1

k

( ),

u i

i

0

k

0,1, 2, ...

( )

u i

1,

i

0,1,

unbounded output

i 0

( )

h i

 

Proof of Necessity (Only if)

Proof of Sufficiency (If)

• Assume the system is BIBO stable but the impulse response is not absolutely summable.

 

• Assume an absolutely summable impulse

 

• Unbounded output with the bounded input (contradiction)

(

uk

i

)

1

,

1

,

()

h i

0

()

h i

0

response and show that the system is BIBO stable.

• Use the input bound b u in the convolution summation

(

y k

)

k

i 0

( )

h i

u

(

k i

)

(

yk

)

k

( ),

hi

 

( ) 

yk

as

k



b

u

k

( )

h i

   k

,

0

 

i 0

13

i 0

14

Theorem 4.3: BIBO Stability

 

Proof of Necessity

 

A discrete-time linear system is BIBO

 

(

h k

)

n

A p

i

k

i

,

k

0,1, 2,

...

Z



 

H

(

z

)

n

A

i

z

stable if and only if the poles of its transfer function lie inside the unit circle.

 

i

1

i

1

z p

i

• Impulse response is bounded if the poles of the transfer function are in the closed unit disc and decays exponentially if the poles are in the open unit disc.

• Systems with a bounded impulse response that does not decay exponentially are not BIBO stable.

Proof of Sufficiency

Assume exponentially decaying impulse response (i.e. poles inside the unit circle). A r is the coefficient of largest magnitude

|p s | < 1 is the largest pole magnitude. The impulse response is bounded by

(

h k

)

n k   A p i i i  0   h i (
n
k
A p
i
i
i  0
h i
( )

n

i 0

n

i 0

A

i

A

r

p

i

0

k p  nA p i r s 1 i  n A s r 1
k
p
 nA
p
i
r
s
1
i
 n A
s
r
1  p
s

k

,

 

k

0,1, 2,

17

Example 4.2

Investigate the BIBO stability of systems with the impulse response

(

h k

)

 

K

0,

,

k

0  

m

 

elsewhere

where K is a finite constant.

18

Solution

BIBO stable since the impulse response satisfies

i

0

()

hi

m

i

0

()

hi

m

1

K



• Let K= upper bound for any impulse response of finite duration.

• Any FIR system is BIBO stable.

19

Example 4.3

Investigate BIBO stability for Example 4.1

a

)

H

(

z

)

4

z

2

z

2



z

0.1

b

)

H

(

z

)

4

z

0.2

z

0.2



z

0.1

c

)

H

(

z

)

5

z

0.3

z

0.2



z

0.1

d

)

H

(

z

)

8

z

0.2

z

0.1

z

1

20

Solution

After pole-zero cancellation

  • a) BIBO stable, all poles inside unit circle.

  • b) BIBO stable, all poles inside unit circle.

  • c) BIBO stable, all poles inside unit circle.

  • d) Not BIBO stable, a pole on unit circle.

21

Z-plane Stable Pole Locations

Im[z]

Unit Circle STABLE Re[z]
Unit Circle
STABLE
Re[z]

22

MATLAB Stability Determination

Obtain roots of polynomial:

» roots(den) % denominator coeffts. den » zpk(g) %g = transfer function

Stable for roots inside the unit circle.

23

MATLAB: ddamp

Gives the pole locations, and n

Eigenvalue

Magnitude Equiv. Damping

Equiv. Freq. (rad/sec)

+ 0.7428I

  • 0.2306 0.7778

0.1941

12.9441

- 0.7428I

  • 0.2306 0.7778

0.1941

12.9441

-0.6612

0.6612

0.1306

31.6871

Closed-loop transfer function >> H = feedback(gforward, gfeedback, 1)

24

Internal Stability

Theorem 4-4

If all the transfer functions that relate the system inputs (R and D) to the possible

 

The system of Figure 4-3 is internally stable if and only if all the closed-loop

 

system outputs (Y and U) are BIBO stable, then the system is internally stable.

D(z E(z) U(z) C(z) G ZAS (z)
D(z
E(z)
U(z)
C(z)
G ZAS (z)

R(z)

+
+

Y(z)

poles are in the open unit disc.

 

D(z E(z) U(z) C(z) G ZAS (z)
D(z
E(z)
U(z)
C(z)
G ZAS (z)

R(z)

+
+

Y(z)

 

C (

)

z G

ZAS

(

z

)

 

G

ZAS

(

z

)

 

R

(

z

D

(

z

   
 

Y

U

(

(

z

)

z

)

 

1

C

(

)

z G

C

(

z

ZAS

)

(

z

)

1

C

(

)

z G

ZAS

(

z

)

C

(

)

z G

ZAS

(

z

)

)

)

 

1

C

(

)

z G

ZAS

(

z

)

1

C

(

)

z G

ZAS

(

z

)

25

26

 

Proof

 

Theorem 4-5

 

C

(

z

)

N

C

(

z

D

C

(

z

)

)

G

ZAS

(

z

)

N

D

G

(

z

)

G

(

z

)

The system of Figure 4-3 is internally stable if

 

Y

1

N

C

N

G

D

C

N

G

  

R

and only if:

 

U

D

C

D

G

N

C

N

G

N

C

D

G

N

C

N

G

 

  

D

Characteristic polynomial D C D G +N C N G

  U   D C D G  N C N G  
 

1.

The characteristic polynomial 1+ C(z)G ZAS (z)

has no zeros on or outside the unit circle. 2. The loop gain C(z)G ZAS (z) has no pole-zero

 

No zeros on or outside the unit circle the transfer functions are asymptotically

all

cancellation on or outside the unit circle .

stable

stable system is internally stable.

system is internally stable.

 
 

27

28

Example 4-4

Transfer function of isothermal chemical reactor

G

(

s

)

0.5848( 0.3549

s

1)

 

0.1828

s 2

0.8627

1

Determine G ZAS (z) with T=0.1 o Verify that the resulting feedback system is not internally stable with the feedback controller

o

C

(

z

) =

10(

z

0.8149)(

z

0.7655)

 

(

z

1)(

z

1.334)

29

Solution

Discretized process transfer function

)   G s (  0.075997 ( z  1.334)  1  Z
)
G s
(
0.075997 (
z 
1.334)
 1
Z
G
(
z
) = 1 
z
ZAS
s
(
z
0.8149)(
z
0.7655)

30

C (

Solution (Cont.)

 10( z  0.8149)( z  0.7655)  0.075997 ( z  1.334) z G
10(
z
0.8149)(
z
0.7655)
0.075997 (
z 
1.334)
z G
)
(
z
) 
ZAS
(
z
1)(
z
1.334)
(
z
0.8149)(
z
0.7655)

violates condition (ii) of Thm. 4-5: pole at 1.334 cancels in the loop gain

31

Routh-Hurwitz Criterion

  • 1. Transform the inside of the unit circle to the LHP (bilinear transformation).

  • 2. Use the Routh-Hurwitz criterion for the investigation of discrete-time system stability.

  • 1 z 1

w

z

 

w

  • 1 z 1

w

32

Advantages/Disadvantages

Example 4.4

• Find stability conditions for • Easy stability test for low-order polynomials. Difficult for high order
Find stability conditions for
Easy stability test for low-order polynomials.
Difficult for high order z-polynomials.
a)
The first order polynomial
For high order polynomials, use symbolic
manipulation.
a z  a
,
a
0
1
0
1
n
n  1
b)
The second order polynomial
Fz
(
)
az
a
z

...
a
n
n  1
0
1  w
n
n  1
2
z 
1  w
1
 w
az
 ,
az
a
a
 0
  a 
1  w
 a
  a
...
2
10
2
n
n  1
1  w
1
 w
0
33
34

Solution: 1 st order

• Solve for the root. • Stability conditions

a 0 a 1
a
0
a
1

1

35

Solution: 2 nd order

z

1 2 ,
1 2
,

2  a a  4 aa 1 1 0 2 2 a 2
2

a
a
 4
aa
1
1
0
2
2 a
2
  • Stability determination by solving for roots is difficult.

  • Monic polynomial

    • constant term = product of poles

  • For pole magnitudes < 1

    • Necessary stability condition

    • Sufficient for complex conjugate poles

a 0 a 2
a
0
a
2

1

   

a

0

a

2

&

a

0

a

2

36

Bilinear Transformation

 

Comments

a

2

1

1

w

w

2

a

1

1

1

w

w

 

  a

0

Recall: The condition is sufficient for

a

  • 210 2



a

aw

2

a

aw

20

a



210

a

a

Routh-Hurwitz criterion: poles of 2nd order w-polynomial

 

complex conjugate roots and only necessary for real roots.

remain in the LHP iff its coefficients are all positive.

 

For real roots, if the three conditions are

aaa   0

210

substituted in the z-domain characteristic

a

2

a

0

0

polynomial we obtain roots between 1

aaa   0

210

and +1.

Adding the first and third conditions gives the condition

 

obtained earlier

a

2

a

0

0

37

38

Stable Parameter Range for 2nd Order z-polynomial

Jury Test

a

2

  • 1 a

a

1

0

1

a

0

  • 1

a

1

a

0

a 0

 1 1  0 a 0 = 1 stable a 0 = a 1 1
 1
1
 0
a 0 = 1
stable a 0 = a 1 1
a 1  1
 0
1
1
a 1
1
 0

The roots of the polynomial are inside the unit

circle iff

 

()

z

 

n

n 1

0

 

0

 

F

az

n

a

(1)

n 1

z



F(1) > 0

az

1

a

0

,

a

n

 

(2)

(1) n F(1) > 0

 

(3)

| a 0 |

< a n

 

(4)

| b 0 | > | b n 1 |

 

(5)

| c 0 | > | c n 2 |

 

39

(n+1)

| r 0 | >

|

r 2 |

 

40

Jury Table

Row

 

z 0

z 1

z 2

 

z n k

 

z n 1

z n

1

a

0

a

1

 

a 2

a n k

a n 1

a

n

2

a

n

a n 1

a n 2

a

k

a

1

a

0

3

 

b

0

b

1

 

b 2

 

b n k

 

b n 1

 

4

b n 1

b n 2

b n 3

b

k

b

0

5

 

c

0

c

1

c

2

 

c n 2

   

6

c n 2

c n 3

c n 4

c

0

.

.

.

.

 

     

.

.

.

.

.

.

.

.

  • 2 s

n 5

 

0

s

1

s

2

s

3

       
  • 2 s

n 4

3

s

2

s

1

s

0

  • 2 r 0

n 3

 

r 1

 

r 2

         

41

Table Entries

b

k

a

0

a

n

a

n

k

a

k

,

k

01,, ,

n



1

c

k

b

0

b

n

1

b

n k

1

b

k

.

,

.

.

k

0,1,

,

n

2

r

0

s

0

s

3

s

3

s

0

r

1

s

0

s

3

s

2

s

1

r

2

s

0

s

3

s

1

s

2

42

Remarks

  • 1. First row of Jury table: listing of the coefficients of the polynomial F(z) in order of increasing power of z.

  • 2. Number of rows of table 2 n 3 is always odd and the coefficients of each even row are the same as the odd row directly above it with the order of the coefficients reversed.

  • 3. The n + 1 conditions for n +1 coefficients of F(z).

  • 4. Conditions 3 through n +1 calculated using the coefficient of the first column of the Jury table, together with the last coefficient of the last row.

  • 5. The middle coefficient of the last row is never used and need not be calculated.

  • 6. Conditions (1) and (2) are calculated from F(z) directly. If one of the first two conditions is violated, F(z) has roots on or outside the unit circle (no need to construct the Jury table or test the remaining conditions).

43

Remarks (cont.)

7- Condition (3) with a

n

= 1, requires the constant term of

the polynomial to be less than unity in magnitude. The

constant term is simply the product of the roots and must

be smaller than unity for all the roots to be inside the unit

circle.

  • 8 For first and second order systems, the Jury stability conditions reduce to the conditions derived earlier.

9- For higher order systems, applying the Jury test by hand

is laborious and it is preferable to test the stability of a

polynomial F(z) using a CAD package.

10- If the coefficients of the polynomial are functions of

system parameters, the Jury test can be used to obtain

the stable ranges of the system parameters.

44

Example

Solution

 

Test the stability of the polynomial

 

(1) F(1) = 1 + 2.6 0.56 2.05 + 0.0775 +0.35 =1.4175 > 0 (2) (1) 5 F(1) = (1)(1+2.6+0.562.050.0775+0.35) = 0.3825 < 0

54

 

056

.

3

2 05

.

 

2

0 0775

.

 

0 35

.

0

 

Fz() z

2 6

.

z

z

z

z

(3) | 0.35 | < 1 (4) | 0.8775 | > | 0.8325 |

Row

z 0

 

z 1

 

z 2

 

z 3

 

z 4

z 5

(5) | 0.0770 | < | 0.5151 |

1

0.35

0.0775

 

2.05

 

0.56

2.6

1

2

1

2.6

 

0.56

2.05

0.0775

 

0.35

(6) | 0.2593 | < | 0.3472 |

3

0.8775

 

2.5729

0.1575

   

1.854

 

0.8325

   

Conditions (2), (5) & (6) violated .

Condition (2) is sufficient to conclude instability.

 

4

0.8325

1.854

 

0.1575

2.5729

0.8775

                   

No. of conditions violated No. of roots outside the unit circle.

                   

5

0.0770

0.7143

 

0.2693

 

0.5151

 
   

Polynomial has roots on or outside the unit circle.

 

6

0.5151

0.2693

 

0.7143

 

0.0770

7

0.2593

 

0.0837

0.3472

           

F ()

z

 z

  • 07. z 05. z 05. z 08. z 25. 0

 

45

•Root at 2.5 outside the unit circle.

46

Example 4.6

Solution

Find the stable range of the gain K for the unity

feedback digital control system with analog plant

Transfer function for analog subsystem, ADC and DAC

G

ZAS

(

z

) 1

z

1

Z _

-1

(

G s

)

s

 

1

1

Z _

-1

K

 

G

(

s

)

K

  z

 

(

s s

3)

 

 

s 3

   

K

K

1

 

1

Partial fraction expansion

with DAC and ADC if the sampling period is 0.02 s.

Transfer function

G

(

s s

3

)

3

s s 3

19412 . 10

2

K

 

ZAS (

z

)

 

47

z 0 . 9418

 

48

Closed-loop System

Unity feedback, closed-loop characteristic equation

1 G

ZAS (

z

)

0

z 0.9418 + 1.9412 10 2 K = 0

Stability conditions

0.9418 1.9412 10 2 K < 1

0.9418 + 1.9412 10 2 K < 1

Hence, the stable range of K is

3 < K < 100.03

49

Example

Find the stable range of the gain K for the digital

position control system with the analog plant

transfer function

(

G s

)

K

(

s s

10)

and with DAC and ADC if T= 0.05 s.

50

Solution

G

ZAS

(

z

)

1

z

1

Z _

-1

(

G s

)

s

 

1

  z

1

Z _

-1

K

s

2

(

s

10)

 

 

Partial fraction expansion

K

s

2

(

s

10)

0.01 K

10

  s

1

 

s

2

1

s

10  

Transfer function

G

ZAS

(

z

)

1.0653 10

2

K

z

0.8467

z

1

z

0.6065

51

C.L. Characteristic Equation

G

ZAS

(

z

)

1.0653 10

2

K

z

0.8467

z

1

z

0.6065

1 G

ZAS

( z ) 0

z

1



z

0.6065

1.0653 10

2

K

z

0.8467

z

2

1.0653 10

2

K

1.6065

z

0.6065 + 9.02

10

3

K

= 0

52

Imaginary Axis

Stability Testing

Imaginary Axis Stability Testing 1.6065 )z+ 0.6065+9.02 1) F (1)=1+ (1.0653  10 K  1.6065)

1.6065)z+0.6065+9.02

1) F(1)=1+ (1.0653 10 2 K 1.6065) +0.6065 + 9.0210 3 K > 0

K > 0

2) F(1) =1(1.065310 2 K1.6065)+0.6065+9.0210 3 K > 0

K < 1967.582

3) | 0.6065 + 0.00902 K | <1

+ (0.6065 + 0.00902 K ) <1

(0.6065 + 0.00902 K ) <1

178.104 < K < 43.6199

&

The stable range is 0 < K < 43.6199

53

Thm. 4-7 Nyquist Criterion

N =no. of counterclockwise encirclements of the

point (1,0) for a loop gain L(z) when traversing

the stability contour (i.e. N clockwise)

L(z) has P open-loop poles inside the contour.

• Then the system has closed-loop poles outside

the unit circle with given by

Z=(N)+P

(N=ZP)

Corollary: An open-loop stable system is closed-

loop stable iff the Nyquist plot does not encircle

the point (1,0) (i.e. N= 0)

54

Example

4.95 z  4.901  5 ( z) = 10 G ZAS 2 z  1.97
4.95
z 
4.901
 5
( z) = 10
G ZAS
2
z
1.97
z
0.9704
Nyquist Diagram
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1

Real Axis

N =0, no poles | | 1, the system is closed-loop stable.

55

Phase Margin and Gain Margin

Gain Margin: gain perturbation that makes

the system marginally stable.

Phase Margin: negative phase perturbation

that makes the system marginally stable.

+ C(z) R(s) G(z) G(z) 
+
C(z)
R(s)
G(z)
G(z)

56

Nyquist Plot: PM & GM

Nyquist Plot: PM & GM 57 MATLAB Plots and Margins >> nyquist(gd) % Nyquist plot. >>

57

MATLAB Plots and Margins

>> nyquist(gd) % Nyquist plot. >> bode(gd) % Bode plot. >> [gm,pm]=margin(gd) % Find PM & GM >> margin(gd) % GM & PM on Bode plot Nyquist plot: click on plot and select Characteristics All stability margins

58

Example

Nyquist Plot: PM & GM 57 MATLAB Plots and Margins >> nyquist(gd) % Nyquist plot. >>

If an amplifier of gain K = 5 is added to the actuator, how does the value of the gain affect closed-loop stability?

59

Solution

Nyquist Diagram 25 20 15 10 5 0 -5 -10 -15 -20 -25 -5 0 5
Nyquist Diagram
25
20
15
10
5
0
-5
-10
-15
-20
-25
-5
0
5
10
15
20
25
30
35
40
45
50
Imaginary Axis

60

Real Axis

GM & PM

Nyquist Diagram 2 K=1, N=0 stable 1.5 K=5 1 Z    N  
Nyquist Diagram
2
K=1, N=0 stable
1.5
K=5
1
Z 
 N
 P
0.5
System: untitled1
Real: -1.41
Imag: 0.000741
Frequency (rad/sec): -5.33
 2  0
0
-0.5
>> [gm,pm]=margin(gtd)
-1
System: gdt
Real: -0.795
Imag: -0.624
Frequency (rad/sec): 2.59
gm =
3.4817
-1.5
pm =
-2
-5
-4.5
-4
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
Real Axis
Imaginary Axis

37.5426

61