12/26/2020

Digital Control and its Applications

AER 407
Lecture 6
System Stability and Root Locus

System Stability
• Closed-loop system transfer function

• Characteristic equation

• A system in the z-plane is stable if all the roots

of the characteristic equation lie inside the
unit circle.

1
 12/26/2020

System Stability
• Several methods used to check the stability of a
discrete-time system:
1. Factorize D(z) = 0 and find the position of the closed
loop poles.
2. Jury’s stability test.
3. Transform the system into the s-plane and analyze
the system stability using s-plane techniques, such
as frequency response analysis or the Routh–
Hurwitz criterion.
4. Use the root-locus graphical technique in the z-
plane to determine the positions of the system
poles.

Factorizing The Characteristic Equation

• Example:

Check the system stability. Assume that T = 1 s.

, for T=1 s

Unstable

2
 12/26/2020

Factorizing The Characteristic Equation

• Example: For the system given in the previous
Example, find the value of T for which the
system is stable.

The Jury Stability Test

– To construct a table whose elements are based on
the coefficients of P(z).
P( z )  a0 z n  a1 z n 1    an 1 z  an

3
 12/26/2020

The Jury Stability Test

– The first row: the coefficients in P(z) in ascending order
– The second row: the coefficients in P(z) in descending order
– Rows 3 through 2n – 3 are given by the following determinants:
an an 1 k
bk  , k  0,1,2,, n  1
a0 ak 1
bn 1 bn  2 k
ck  , k  0,1,2,, n  2
b0 bk 1

p3 p2  k
qk  , k  0,1,2
p0 pk 1

• The last row in the table consists of three elements.

• The elements in any even-numbered row are simply the reverse of the
immediately preceding odd-numbered row.

The Jury Stability Test

• Stability Criterion by the Jury Test
– A system with the characteristic equation
P( z )  a0 z n  a1 z n 1    an 1 z  an  0, where a0  0
is stable if the following conditions are all satisfied:

1. an  a0
2. P( z )  0
z 1

 0 for n even
3. P( z ) z 1 
  0 for n odd

4. bn 1  b0
cn  2  c0

q 2  q0

4
 12/26/2020

The Jury Stability Test

• Example
P( z )  z 4  1.2 z 3  0.07 z 2  0.3z  0.08  0
First condition is satisfied: a4  a0
Second condition is satisfied: P(1)  1  1.2  0.07  0.3  0.08  0.09  0
Third condition is satisfied:
P(1)  1  1.2  0.07  0.3  0.08  1.89  0, n  4  even
The Jury stability table: (next page)

Fourth condition is satisfied:

b3  0.994  0.204  b0
c2  0.946  0.315  c0
Therefore, the given characteristic equation is stable, or all roots lie inside the
unit circle in the z plane.
P( z )  ( z  0.8)( z  0.5)( z  0.5)( z  0.4)

The Jury Stability Test

5
 12/26/2020

The Jury Stability Test

• Example:
The open loop of the system is
Determine the value of K for which the system is
stable

Thus, the system is stable for 0.285 < K < 2

Root-Locus
• The root-locus method developed for continuous-time systems can be
extended to discrete-time systems without modifications, except that the
stability boundary is changed from the jw axis in the s plane to the unit
circle in the z plane.
• Characteristic equation is 1+G(z)H(z)=0 which is of exactly the same
form as the equation for root-locus analysis in the s plane. However, the
pole locations for closed-loop systems in the z plane must be interpreted
differently from those in the s plane

6
 12/26/2020

Root-Locus
• Angle and Magnitude Conditions
1  G( z) H ( z)  0 and 1  GH ( z )  0 1  F ( z)  0
F ( z)  G( z) H ( z) or F ( z )  GH ( z )

F ( z )  1 Angle condition: F ( z )  180(2k  1), k  0,1,2, 

Magnitude condition: F ( z)  1

– The values of z that fulfill both the angle and the

magnitude conditions are the roots of the
characteristic equation, or the closed-loop poles.

Root-Locus
• General Procedure for Constructing Root Loci
1. Obtain the characteristic equation and rearrange this equation
so that the parameter of interest, K, appears as the multiplying
factor in the form.
1  F ( z)  0
K ( z  z1 )( z  z2 ) ( z  zm )
1 0
( z  p1 )( z  p2 ) ( z  pn )

2. Find the starting points and terminating points of the root loci.
Find also the number of separate branches of the root loci.
• The point on the root loci corresponding to K =0 are open-loop poles
and those corresponding to K= are open-loop zeros.
• A root-locus plot will have as many branches as there are roots of the
characteristic equation.

7
 12/26/2020

Root-Locus
• General Procedure for Constructing Root Loci. (cont.)
3. Determine the root loci on the real axis.
• Root loci on the real axis are determined by open-loop poles and
zeros lying on it.
• In constructing the root loci on the real axis, choose a test point on it.
If the total number of real poles and zeros to the right of this test
point is odd, then this point lies on a root locus. The root locus and
its complement form alternate segments along the real axis.

4. Determine the asymptotes of the root loci.

 180(2 N  1) n=number of poles of F(z)
Angle of asymptote  , N  0,1,2,...
nm m=number of zeros of F(z)

• The intersection of the asymptotes and the real axis.

( p1  p2    pn )  ( z1  z 2    z m )
 a  
nm

Root-Locus
• General Procedure for Constructing Root Loci. (cont.)
5. Find the breakaway and break-in points.
• Because of the conjugate symmetry of the root loci, the
breakaway points and break-in points either lie on the real axis or
occur in conjugate complex pairs.
KB( z ) A( z )
1  F ( z)  0 1 0 K 
A( z ) B( z )
dK A( z ) B( z )  A( z ) B( z )
 0
dz B 2 ( z)
• If the value of K corresponding to a root z=z0 of dK/dz=0 is
positive, point z=z0 is an actual breakaway or break-in point.
• Since K is assumed to be non negative, if the value of K is
negative, then point z=z0 is neither a breakaway nor a break-in
point.

8
 12/26/2020

Root-Locus
• General Procedure for Constructing Root Loci. (cont.)
6. Determine the angle of departure (or angle of arrival) of the root loci
from the complex poles (or at the complex zeros).
• The angle of departure (or angle of arrival) of the root locus from a
complex pole (or at a complex zero) can be found by subtracting from
180 the sum of all the angles of lines from all other poles and zeros to
the complex pole in question.

Root-Locus
• General Procedure for Constructing Root Loci. (cont.)
7. Find the points where the root loci cross the imaginary axis.
• The points where the root loci intersect the imaginary axis can be
found by setting z=jv in the characteristic equation, equating both
the real part and the imaginary part to zero, and solving for v and K.
The values of v and K found give the location at which the root loci
cross the imaginary axis and the value of the corresponding gain K,
respectively.

8. Any point on the root loci is a possible closed-loop pole.

F ( z)  1
( z  z1 )( z  z2 )  ( z  zm ) 1

( z  p1 )( z  p2 )  ( z  pn ) K
• A particular point will be a closed-loop pole when the value of gain K
satisfies the magnitude condition.

9
 12/26/2020

Root-Locus
• Root-Locus Diagrams of Digital Control Systems
– The effects of gain K and sampling period T on the relative stability of
the closed-loop control system

K z
GD ( z )  K
1  z 1 z 1
1  e Ts 1   1 
 
Z Gh ( s )G p ( s )     (1  z )Z 
s  1
1


1 1 
  (1  z )Z  s  s  1
1

 s  s ( s 1)   
T
z 1  z z  1 e
   
z  z  1 z  e T  z e
T

Kz 1  e T 
Char. Eq. : 1  Kz 1  e
T

 
G ( z )  GD ( z )Z Gh ( s)G p ( s) 
z  1 z  e T z  1z  e
T
0


Root-Locus
• Root-Locus Diagrams of Digital Control Systems
1. Sampling period T =0.5 sec:
0.3935Kz
G( z )  The closed-loop poles corresponding to K=2:
z  1z  0.6065
z1  0.4098  j 0.6623 and
  z  1 z  0.6065 z2  0.4098  j 0.6623
K
0.3935z
dK z 2  0.6065
 0
dz 0.3935z 2

z 2  0.6065
z  0.7788 and z  0.7788
0.3935z 1

( z  1)( z  0.6065) K

The critical gain KC: (z=-1)

KC =8.165

10
 12/26/2020

Root-Locus
• Root-Locus Diagrams of Digital Control Systems (cont.)
2. Sampling period T =1.0 sec:
0.6321Kz
G( z )  The closed-loop poles corresponding to K=2:
z  1z  0.3679
z1  0.05185  j 0.6043 and
The breakaway and break-in points
z 2  0.05185  j 0.6043
z  0.6045 and z  0.6065

The critical gain KC: KC =4.328

Root-Locus
• Root-Locus Diagrams of Digital Control Systems (cont.)
3. Sampling period T =2.0 sec:
0.8647Kz
G( z )  The closed-loop poles corresponding to K=2:
z  1z  0.1353
z1  0.2971 j 0.2169 and
The breakaway and break-in points
z 2  0.2971 j 0.2169
z  0.3678 and z  0.3678

The critical gain KC: KC =2.626

11
 12/26/2020

Root-Locus
• Effects of Sampling Period T on Transient Response
Characteristics
– For a given value of gain K, increasing the sampling period T will
make the discrete-time control system less stable and eventually will
make it unstable.
• Making T shorter and shorter tends to make the system behave more
like the continuous-time system.

Root-Locus
• Effects of Sampling Period T on Transient Response
Characteristics (cont.)
– For T=1.0 sec and K=2
1.2642z 1 1
C ( z) 
1  0.1037z 1  0.3679z 2 1  z 1

– For T=2.0 sec and K=2

1.7294z 1 1
C( z) 
1  0.5941z 1  0.1353z 2 1  z 1

12
 12/26/2020

Design based on the Root-Locus

Method
• Example

– Design a digital controller such that the dominant closed-

loop poles have a damping ratio() of 0.5 and settling
time(ts) of 2 sec. Assume T=0.2 sec.
– Obtain the unit-step response of the designed digital
system.
– Obtain Kv of the system.

Design based on the Root-Locus

Method
• Example (cont.)
– For the standard second-order system

4 4
ts   2  n  4
 n 0.5 n

 d   n 1   2  4 1  0.52  3.464

– Location of the desired dominant closed-loop poles

z  e Tn  e 0.4  0.6703

z  Td  0.6927rad  39.69
z  0.670339.69  0.5158  j 0.4281

13
 12/26/2020

Design based on the Root-Locus

Method
• Example (cont.)
– The pulse transfer function of ZOH and the plant: G(z)
1  e 0.2 s 1  1  1  0.1758( z  0.8760)
G( z)  Z    (1  z ) Z  2 
 s s( s  2)   s ( s  2)  ( z  1)( z  0.6703)

• poles : z=1, z=0.6703

• zero : z=-0.8760
– The sum of the angle contribution at P
17.10  135.52  109.84  231.26
• The angle deficiency is –231.26+180 = -51.26
and needs to be provided by the controller.

109.84 135.52
17.10

Design based on the Root-Locus

Method
• Example (cont.)
– The controller may be assumed to be
z 
GD ( z )  K
z

• If we decide to cancel the pole at z=0.6703 by the zero of GD(z) at z=-

, then the pole of the controller can be determined as z=0.2543.
z  0.6703
GD ( z )  K
z  0.2543
• The open-loop
z  0.6703 0.01758( z  0.8760)
GD ( z )G ( z )  K
z  0.2543 ( z  1)( z  0.6703)
0.01758( z  0.8760)
K
( z  0.2543)( z  1)

• The gain K can be determined

17.10 109.84 135.52
GD ( z )G ( z ) z 0.5158 j 0.4281  1
K  12.67

14
 12/26/2020

Design based on the Root-Locus

Method
• Example (cont.)
– The designed digital controller
z  0.6703
GD ( z )  12.67
z  0.2543

0.01758( z  0.8760) 0.2227( z  0.8760)

GD ( z )G( z )  12.67 
( z  0.2543)( z  1) ( z  0.2543)( z  1)

• The closed-loop pulse transfer function

C ( z) GD ( z )G ( z ) 0.2227z  0.1951
 
R( z ) 1  GD ( z )G( z ) z 2  1.0316z  0.4494

• The response to the unit-step input

0.2227z  0.1951 1 0.2227z 1  0.1951z 2 1
C ( z)  
z 2  1.0316z  0.4494 1  z 1 1  1.0316z 1  0.4494z 2 1  z 1

Design based on the Root-Locus

Method
• Example (cont.)
– The static velocity error constant KV
1  z 1  1  z 1 0.2227( z  0.8760) 
KV  lim  GD ( z )G ( z )  lim    2.801
z 1
 T  z 1
 0.2 ( z  0.2543)( z  1) 
• If it is required to have a large value of KV, then we may include a
lag compensator.
( z  0.94)
3
( z  0.98) z 1

– The pole and zero for the compensator lie on a finite number of allocable
discrete points. – the pole and the zero are very closed to each other and
do not significantly change the root locus near the dominant closed-loop
poles.
– Even the designed system is of the third order, it acts as a
second-order system, since one pole of the plant has been
canceled by the zero of the controller.

15
 12/26/2020

Thank You

16