Vous êtes sur la page 1sur 66

EEE2502 CONTROL ENGINEERING III

Course Content
State variable methods: state space mode representation, similarity transformations, state transi-
tion, flow graphs and canonical forms; geometrical interpretation of eigenvalues and eigen-vectors;
time solution by transition matrix and inverse. Laplace techniques: canonical forms in relation to
controllability and observability; the observer, simple pole placement design technique.
Discrete systems: sampling, data extrapolators and spectral characteristics; design of Z-transform,
impulse in variance and hold equivalent; block manipulation and pulse transfer functions; Z-
transform inversion; stability and damping; discrete compensator realization design examples; bi-
linear transformation; root space. Frequency domain identification using least squares.

References
1. Modern Control Engineering, Kastuhiko Ogata

2. Discrete Time Control Systems, Kastuhiko Ogata

3. Advanced Control Engineering, Roland S. Burns

4. Modern Control Engineering , P. N. Paraskevopoulos

1
Contents
1 STATE VARIABLE METHODS 3
1 State space representation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 State Equations from Transfer Functions . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Similarity Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 Eigen Values and Eigen Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5 Repeated Eigen values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
6 Laplace Transform Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7 Transfer Function from state space equations . . . . . . . . . . . . . . . . . . . . . . 20
8 Time Solution of state equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
9 Controllability and Observability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
10 Pole Placement Design Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
11 State Observer Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

2 DISCRETE SYSTEMS 46
1 Sampled Data Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2 The Z-Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3 Inverse Z Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

2
1 STATE VARIABLE METHODS

1 State space representation

State space approach is a generalized time-domain method of modelling, analysing and designing
a wide range of control systems.
The approach can deal with

1. Multiple input, multiple output (MIMO) systems or multivariable systems

2. Nonlinear and time invariant systems

3. Alternative controller design approaches

The analysis require

1. Input variables

2. Output variables

3. State variables

Definitions

State: State of dynamic system is the smallest set of variables (state variables) such that knowl-
edge of the variables at time t = to plus knowledge of the inputs for t ≥ to , this information
determines the behaviour of the system for any time t ≥ to .

State variables: These are variables making up the smallest set of variables that determine the
state of the dynamic system.Let a dynamic system have n variables x1 , x2 , . . . ., xn to describe the
behavior of the system while the input is given fort ≥ to and the initial state at t = to is specified.
If the future state of the system is determined, then such variables are a set of state variables.

State vector: The behavior of a given system is described by n state variables and can be
considered to be n components of a vector x. Such a vector is called a state vector. A state is thus
a vector that determines uniquely the system state x (t) for anyt ≥ to once the state at t = to is
given and the input for t ≥ to is specified.

State space: The n dimensional space whose coordinate axis consists of x1 , x2 , . . . ., xn axis is
called state space. Any state can be represented by a point in the state space.
Consider the dynamic system shown below
The above dynamic system has MIMO.

3
The state of the system is described by a set of first order differential equations in terms of the
state variables x1 , x2 , . . . ., xn and input variables u1 , u2 , . . . ., ur as

ẋ1 (t) = f1 (x1 , x2 , . . . , xn ; u1 , u2 , . . . , ur ; t)

ẋ2 (t) = f2 (x1 , x2 , . . . ., xn ; u1 , u2 , . . . ., ur ; t) (1)

ẋn (t) = (fn x1 , x2 , . . . , xn ; u1 , u2 , . . . , ur ; t)

The output of the system y1 (t) , y2 (t) ,. . .,ym (t) may be a function of input variables, state
variables and time. This may be described as

y1 (t) = (g1 x1 , x2 , . . . , xn ; u1 , u2 , . . . , ur ; t)
y2 (t) = g2 (x1 , x2 , . . . ., xn ; u1 , u2 , . . . ., ur ; t) (2)

ym (t) = gm (x1 , x2 , . . . ., xn ; u1 , u2 , . . . ., ur ; t)

Definitions of u (t) , x (t) and y (t) would be


 
x1 (t)
 x2 (t) 
 
x (t) =  . 

 state vector
 .. 
xn (t)

 
u1 (t)
u2 (t)
 
y (t) =  . 

 input vector
 .. 
ur (t)
 
y1 (t)
 y2 (t) 
 
y (t) = 
 ..  output vector

 . 
ym (t)

Using the above vectors, equations (1) and (2) become

ẋt = f (x, u, t) state equation (3)

y (t) = g (x, u, t) output equation (4)

4
Time invariant systems
If the vector functions of f and g do not involve time, then the system is said to be a time invariant
system
Then
ẋt (t) = f (x, u) (5)

y (t) = g (x, u) (6)

If equations 5 and 6 are linear then

ẋt = Ax (t) + Bu (t) (7)

y (t) = Cx (t) + Du (t) (8)

where A, B, C and D are constant matrices

Time varying systems


If the system is time varying, then equations 3 and 4 result to

ẋ (t) = A (t) x (t) + B (t) u (t) (9)

y (t) = C (t) x (t) + D (t) u (t) (10)

where
A (t) – state matrix
B (t) – input matrix
C (t) – output matrix
D (t) – direct transition matrix
Example 1.1
Obtain the state space representation of the system described by

d3 y d2 y dy
3
+ 6 2 + 11 + 6y = 6u
dt dt dt
where y is the output andu is the input to the system
Solution
Knowledge of y (0) , ẏ (0)and ÿ(0) together with the input u (t) for t ≥ 0 determines completely the
future of the system. Thus if y (t) , ẏ (t)and ÿ(t)are a set of the state variables then

5
Defining the state variables as

x1 = y

x2 = ẏ

x3 = ÿ

Then

ẋ1 = ẏ = x2

x2 = ẏ = x3

x3 = ÿ = −6x1 − 11x2 − 6x3 + 6u

By use of vector matrix notation, then the 3 first order differential equations can be combined into
one as       
ẋ1 0 1 x1
0 0
ẋ =
 2  0 0 1  x2  + 0 u
      

ẋ3 −6 −11 −6 x3 6

The output y is given by

 
h i x1
y= 1 0 0 x2 
 

x3

Example 1.2
Write the state variable formulation of the parallel RLC network shown below

input is i (t) = I sin ωt


Solution
Applying KCL at node A then
i = iR + iC + iL

6
ˆ
v (t) dv (t) 1
I sin ωt = +C + v (t) dt
R dt L

Differentiating with respect to t and rearranging yields

d2 v 1 dv 1 ω
2
+ + v = I cos ωt
dt RC dt LC C
Choosing
v (t) = x1 (t)

ẋ1 (t) = x2

1 1 ω
ẋ2 (t) = − x1 − x2 + I cos ωt
LC RC C
The vector-matrix differential form of the state equation can be written as
" # " #" # " #
ẋ1 0 1 x1 0
= 1 1
+ ω
ẋ2 − LC − RC x2 CI cos ωt

and the output " #


h i x
1
v= 1 0
x2

NB
Usually in the circuit problem, the current through the inductor and voltage across the capacitor
are chosen as the state variables

2 State Equations from Transfer Functions

Consider a system defined by

y (n) + a1 y (n−1) + · · · + a(n−1) ẏ + an y = b0 un + b1 un−1 + · · · + bn−1 u̇ + bn u (1)

where u is the input and y is the output


The equation can be written as

Y (s) b0 sn + b1 sn−1 + · · · + bn−1 s + bn


= n (2)
U (s) s + a1 sn−1 + · · · + an−1 s + an

The state space representation of the system defined by equation (11) and (12) can be presented
in controllable canonical form, observable canonical form and diagonal canonical form

7
Controllable canonical form

The controllable canonical form of the state space representation is given by


     
ẋ1 0 1 0 ··· ··· 0 x1 0
 ẋ2   0 0 1 ··· ··· 0   x2  0
      
 ..   .. .. .. ..   ..   .. 
      
 .   . . . .   .  .
 . =  .  + . u (3)
      
 ..   .. .. .. ..   ..   .. 
   . . . . 
   
ẋn−1   0 0 0 ··· ··· 1  xn−1  0
      

ẋn −an −an−1 −an−2 ··· ··· −a1 xn 1

and

 
x1
 x2 
 
. 
 
h  .. 
i
y = bn − an b0 bn−1 − an−1 b0 ··· ··· b1 − a1 b0 
 .  + b0 u
 (4)
 .. 
 
xn−1 
 

xn

Example 1.3
Consider the system given by

Y (s) s+3
= 2
U (s) s + 3s + 2

Obtain the controllable canonical form of the state space representation


Solution
n = 2 , b0 = 0, b1 = 1, b2 = 3 ,a1 = 1, a2 = 2
" # " #" # " #
ẋ1 0 1 x1 0
= + u
ẋ2 −a2 −a1 x2 1
" #" # " #
0 1 x1 0
= + u
−2 −3 x2 1
" #
h i x
1
y = b2 − a2 b0 b1 − a1 b0 + b0 u
x2
" #
h i x
1
= 3 1
x2

8
Observable canonical form

For the transfer function

Y (s) b0 sn + b1 sn−1 + · · · + bn−1 s + bn


= n
U (s) s + a1 sn−1 + · · · + an−1 s + an

The observable canonical form of the state space representation is given by

     
ẋ1 0 0 ··· ··· 0 −an x1 bn − an b0

 2  1 0 ··· ··· 0 −an−1   x2  bn−1 − an−1 b0 
      
 ..   .. .. ..   ..   ..
      

 .  . . .   .   . 
u
 .  = .
    + (5)
 ..   .. .. ..   .  
. .
.

   . .  .  
    . 

ẋn−1  0 0 ··· ··· 0 −a2  xn−1   b2 − a2 b0 
      

ẋn 0 0 ··· ··· 1 −a1 xn b1 − a1 b0

and

 
x1
 x2 
 
. 
 
h  .. 
i
y= 0 0 ··· ··· 0  .  + b0 u
1   (6)
 .. 
 
xn−1 
 

xn

Example 1.4
Obtain the observable canonical form of the state space representation for the system with the
transfer function

Y (s) s+3
= 2
U (s) s + 3s + 2

Solution
n = 2 , b0 = 0 ,b1 = 1 ,b2 = 3 ,a1 = 1 , a2 = 2
" # " #" # " #
ẋ1 0 −a2 x1 b2 − a2 b0
= + u
ẋ2 1 −a1 x2 b1 − a1 b0
" #" # " #
0 −2 x1 3
= + u
1 −3 x2 1
" #
h i x
1
y= 0 1 + b0 u
x2

9
" #
h i x
1
= 0 1
x2

Diagonal canonical form

If the denominator polynomial has distinct roots then the transfer function can be written as

Y (s) b0 sn + b1 sn−1 + · · · + bn−1 s + bn


= (7)
U (s) (s + p1 ) (s + p2 ) · · · (s + pn−1 ) (s + pn )

c1 c2 cn−1 cn
= b0 + + + + (8)
(s + p1 ) (s + p2 ) (s + pn−1 ) (s + pn )

the diagonal canonical form of the state space representation of the system is given by

    
ẋ1 −p1 0 ··· ··· 0 0 x1 1
 ẋ2   0 −p1 ··· ··· 0   x2  1
      
 ..   .. .. .. ..   ..   .. 
      
 .   . . . .  .   + . u
   
 . = . (9)
  
 ..   .. .. .. ..   ..  
   .. 
   . . .  .   .


 n−1   0 0 0 ··· −pn−1 0  xn−1  1
      

ẋn 0 0 −an−2 ··· ··· −pn xn 1

and

 
x1
 x2 
 
. 
 
h  .. 
i
y = c1 c2 ··· ··· cn−1  .  + b0 u
cn   (10)
 .. 
 
x
 
 n−1 
xn

Example1.5
Obtain the diagonal canonical form of the state space representation for the system with the transfer
function

Y (s) s+3
= 2
U (s) s + 3s + 2

Solution

s+3 s+3 c1 c2
= = +
s2 + 3s + 2 (s + 1) (s + 2) (s + 1) (s + 2)

10
b0 = 0 , c1 = 2 , c2 = −1 , p1 = 1 , p2 = 2
" # " #" # " #
ẋ1 −p1 0 x1 1
= + u
ẋ2 0 −p2 x2 1
" #" # " #
−1 0 x1 1
= + u
0 −2 x2 1
" #
h i x
1
y = c1 c2 + b0 u
x2
" #
h i x
1
= 2 −1
x2

Jordan Canonical Form

If the denominator polynomial involves multiple roots then the state space representation can be
written in Jordan canonical form.
For example, if the pi ‘s are different from one another except that the first three are equal i.e.
p1 = p2 = p3 , then the factored form of the transfer function becomes

Y (s) b0 sn + b1 sn−1 + · · · + bn−1 s + bn


= 3 (11)
U (s) (s + p1 ) (s + p4 ) · · · (s + pn )

The partial fraction expansion of the equation becomes

Y (s) c1 c2 c3 c4 cn
U (s)
= b0 + 3 + 2 + (s + p ) + (s + p ) + · · · + (s + p ) (12)
(s + p1 ) (s + p1 ) 1 4 n

The state space representation of the system in Jordan canonical form becomes

 
 
ẋ1 −p1 1 0 0 0    
 x1 0
−p1

   0 1 0
 ẋ2  
   
  x2  0
   ..    
 ẋ3   0 0 −p1 0 .   x  1
  3  
 =   +  u (13)
  
 ẋ4   0 ..   x4  1
 .   −p4 .    
 .   .    
 .   .. ··· 0
   
 
ẋn xn 1
0 ··· ··· ··· ··· −pn

and

11
 
x1
 
 x2 
h i 
 .. 
y = c1 c2 ··· ··· cn  . 

 + b0 u (14)
 . 
 .. 
 
xn

The characteristic equation of a given system remains invariant under different forms of state
variable representation. This can be true for the transfer function also. The choice of the states is
not unique
Example 1.6
Consider the system
...
y + 6ÿ + 11ẏ + 6y = 6u

where y is the output and u is the input. Obtain the state space representation of the system
Solution I
Let
x1 = y

x2 = ẏ

x3 = ÿ

ẋ1 = x2

ẋ2 = x3

ẋ3 = −6x1 − 11x2 − 6x3 + 6u

These can be written in the vector matrix differential equation form as

      
ẋ1 0 1 x1 0 0
ẋ2  =  0 0 1  x2  + 0 u
      

ẋ3 −6 −11 −6 x3 6
 
h i x1
y= 1 0 0 x2 
 

x3

The characteristic equation becomes

|SI − A| = 0

12
s3 + 6s2 + 11s + 6 = 0

Solution II

...
y + 6ÿ + 11ẏ + 6y = 6u

The transfer function is obtained as

Y (s) 6
= 3 2
U (s) s + 6s + 11s + 6

6
=
(s + 1) (s + 2) (s + 3)

By partial fraction expansion

Y (s) 3 −6 3
= + +
U (s) (s + 1) (s + 2) (s + 3)

3 −6 3
Y (s) = U (s) + U (s) + U (s)
(s + 1) (s + 2) (s + 3)

Defining
Y (s) = X1 (s) + X2 (s) + X3 (s)

where

3
X1 (s) = U (s) ẋ1 = −x1 + 3u
(s + 1)

−6
X2 (s) = U (s) ẋ2 = −2x2 − 6u
(s + 2)

3
X1 (s) = U (s) ẋ3 = −3x3 + 3u
(s + 3)

The state space representation becomes

      
ẋ1 1 0 0 x1 3
ẋ2  = 0 −2 0  x2  + −6 u
      

ẋ3 0 0 −3 x3 3
 
h i x1
y= 1 1 1 x2 
 

x3

13
|SI − A| = 0

s3 + 6s2 + 11s + 6 = 0

3 Similarity Transformation
It has been stated that the choice of states is not unique for a given system. Suppose that there
exists a set of state variables

h iT
X = x1 x2 ··· xn (1)

We may take another set of state variables

h iT
Z = z1 z ··· zn (2)

So that a linear or similarity transformation exists


Let X = P Z i.e
Z = P −1 X (3)

where P is a non-singular transformational matrix


Differentiating equation (27) yields

Ż = P −1 Ẋ (4)

Using the general state space equations

Ẋ = Ax + Bu

y = Cx

Then equation (28) becomes

Ż = P −1 Ax + P −1 Bu (5)

From equation (27) X = P Z


Equation (29) becomes

Ż = P −1 AP z + P −1 Bu (6)

and
y = Cx = CP z (7)

14
Equations (30) and (31) can be written as

Ż = Âz + B̂u (8)

y = Ĉx (9)

Where  = P −1 AP, B̂ = P −1 B and Ĉ = CP


Hence similarity transformation the transformed system can be represented in the vector-matrix
differential form as
Ż = Âz + B̂u

and the output as


y = Ĉx

NB

1. The characteristic equations and hence the Eigen values of A and Âare invariant under simi-
larity transformation

2. The transfer function remains invariant under similarity transformation

4 Eigen Values and Eigen Vectors

Given the matrix equation

Ax = λx (1)

The values of the scalar λ for which non trivial solutions exist are called Eigen values and the
corresponding solutions x = 0 are called Eigen vectors
Equation (34) can be written in the form
(λI − Ax) = 0 whereI is the identity matrix
|λI − Ax| = 0 is the characteristic equation of A
The roots of the characteristic equation are called Eigen values of the matrix A
Corresponding to each Eigen value is a non-zero solution of x = expi This is called the Eigen vector
of A corresponding to λi
Example 1.7
" #
4 1
Determine the Eigen values and Eigen vectors of Ax = λx where A =
3 2
Solution
|λI − A| = 0

15
(λ − 4) (λ − 2) − 3 = 0

λ1 = 5, λ2 = 1

For λ1 = 5 , Ax = λx becomes " #" # " #


4 1 x1 x1
=5
3 2 x2 x2

4x1 + x2 = 5x1

3x1 + 2x2 = 5x2

x1 = x2

Eigen vector corresponding toλ1 = 5 is


" # " #
x1 1
exp1 = =
x2 1
simplest form
Forλ2 = 1 , Ax = λx becomes
" #" # " #
4 1 x1 x1
=1
3 2 x2 x2

4x1 + x2 = x1

3x1 + 2x2 = x2

x1 = −3x2

Eigen vector corresponding to λ2 = 1 is


" # " #
x1 1
exp2 = =
x2 −3

simplest form

16
We obtain the transformation matrix P from the Eigen vectors as follows
" #
h i 1 1
P = exp1 exp2 =
1 −3

The matrix = P −1 AP is then obtained

" #−1 " #" # " #


1 1 4 1 1 1 5 0
 = P −1 AP = =
1 −3 3 2 1 −3 0 1

Under similarity transformation the characteristic equation does not change i.e.


|λI − A| = λI − Â = λ2 + λs + 5

5 Repeated Eigen values

In some cases, the matrix A will have repeated Eigen values. The Eigen vectors are evaluated as
follows
Example 1.8
 
3 −3 2
Determine the Eigen values and Eigen vectors of Ax = λx where A = −1 5 −2
 

−1 3 0
Solution
|λI − A| = 0


λ − 3 3 −2

1 λ−5 2 =0


1 −3 λ

λ3 − 8λ2 + 20λ − 16 = 0

λ1 = 4 λ2 = λ3 = 2

For λ1 = 4

    
3 −3 2 x1 x1
−1 5 −2 x2  = 4 x2 
    

−1 3 0 x3 x3

This yield

17
x2 = −x1

and
x3 = −x1

and the simplest form of the corresponding Eigen vector as

 
1
exp1 = −1
 

−1

For λ2 = λ3 = 2

    
3 −3 2 x1 x1
−1 5 −2 x2  = 2 x2 
    

−1 3 0 x3 x3

The simultaneous equations obtained are

x1 − 3x2 + 2x3 = 0

We let x2 = α and x3 = βwhere are α and β constants


The above equation becomes

x1 − 3α + 2β = 0

x1 = 3α − 2β

and the resulting Eigen vector as


   
3 −2
exp = α 1 + β  0 
   

0 1

 
3
For β = 0and α = 1 , exp2 = 1
 

0
 
−2
Forβ = 1 and α = 0 ,exp3 =  0 
 

18
The transformation matrix P is obtained as
 
1 3 −2
P = −1 1 0
 

−1 0 1

The matrix  = P −1 AP is then obtained as

 −1   
1 3 −2 3 −3 2 1 3 −2
 = P −1 AP = −1 1 0 −1 5 −2 −1 1 0
    

−1 0 1 −1 3 0 −1 0 1
 
4 0 0
= 0 2 0
 

0 0 2

Checking the characteristic equation


|λI − A| = λI − Â = λ3 − 8λ2 + 20λ − 16 = 0

6 Laplace Transform Technique

Recall I
The Laplace transform of a function f (t) is defined as the integral
ˆ ∞
L {f (t)} = exp−st f (t) dt (1)
0

Example 1.9
Obtain the Laplace transform of f (t) = expat
Solution
ˆ ∞
L {f (t)} = exp−st expat dt
0

1
=
s−a
Recall II

n 0 o
L f (t) = sL {f (t)} − f (0) (2)

0
L f ” (t) = s2 L {f (t)} − sf (0) − f (0)

(3)

19
and so on for Laplace transform of higher derivatives(you should be able to prove this with a lot of
ease)
Example 1.10
Use the Laplace transform of second derivative to derive

s
L {cos (at)} =
s2 + a2
Solution
Let
f (t) = cos (at)

0
f (t) = −a sin at

and

f ” (t) = −a2 cos (at)

0
f (0) = 0

f (0) = 1

Using equation (37)

0
L f ” (t) = s2 L {f (t)} − sf (0) − f (0)


L −a2 cos (at) = s2 L {cos (at)} − s − 0




s2 + a2 L {cos (at)} = s


s
L {cos (at)} =
s2 + a2

7 Transfer Function from state space equations


Y (s)
The transfer function is given by G (s) = U (s)

The general state space representation of any given systems is

ẋ = Ax + Bu (1)

20
y = Cx + Du (2)

Taking the Laplace transforms of (38) and (39)

sX (s) − X (0) = AX (s) + BU (s) (3)

Y (s) = CX (s) + DU (s) (4)

Assuming that the initial conditions of the system are zero i.e. X (0) = 0 and rearranging (40) then

−1
X (s) = (SI − A) BU (s) (5)

Substituting (42) into (41) yields


 
−1
Y (s) = C (SI − A) B + D U (s) (6)

and the transfer function is obtained as

Y (s)  −1

G (s) = = C (SI − A) B + D (7)
U (s)

Example 1.11
Obtain the transfer function of the system whose state space representation is given as
" #" # " #
0 1 x1 0
ẋ = + u
−2 −3 x2 1
" #
h i x
1
y= 1 0
x2

Solution

Y (s)  −1

G (s) = = C (SI − A) B + D
U (s)
where
" # " #
0 1 0 h i
A= , B= , C= 1 0 and D = 0
−2 −3 1
" #
−1 1 s+3 1
(SI − A) = 3
s + 3s + 2 −2 s

21
" #" #
  1 h i s+3 1 0
−1
C (SI − A) B + D = 1 0
s3 + 3s + 2 −2 s 1

1
=
s3 + 3s + 2

8 Time Solution of state equations

Consider the homogenous case


ẋ (t) = ax (t) (1)

Approach A
Assume a solution of the form

x (t) = b0 + b1 t + b2 t2 + · · · + bk tk (2)

Substituting (46) into (45) yields

ẋ = b1 + 2b2 t + · · · + kbk tk−1 = ab0 + b1 t + b2 t2 + · · · + bk tk


 
(3)

If the assumed solution is to be a true solution, equation (47) must hold true for any value of t
This implies that

b1 = ab0

1 2
b2 = a b0
2

1 k
bk = a b0
k!
The value of b0 is determined by substituting t = 0 into equation (46) i.e.

X (0) = b0

The solution of x (t) can be written as


 
1 2 2 1 k k
x (t) = 1 + at + a t + · · · + a t x (0)
2 k!

But
1 1
expat = 1 + at + a2 t2 + · · · + ak tk
2 k!
Therefore

22
x (t) = expat x (0) (4)

Approach B
Using Laplace transform of the homogeneous equation (45)

sX (s) − X (0) = aX (s)

1
X (s) = X (0)
s−a
Taking the inverse Laplace transform

x (t) = expat x (0)

State Transition Matrix

The approach to the solution of homogeneous scalar equation can be extended to the solution of
homogeneous state equation

ẋ (t) = Ax (t) (5)

Taking Laplace transform

sX (s) − X (0) = AX (s)

−1
X (s) = (sI − A) X (0)

Taking the inverse Laplace transform

−1
x (t) = L−1 (sI − A) x (0) (6)

x (t) = expAt x (0) (7)

Where

−1
expAt = L−1 (sI − A)

expAt is called the state transition matrix and contains all information about the free motions of
the system described by (49)
Example 1.12

23
Obtain the state transition matrix of the following state equation
" # " #" #
ẋ1 0 1 x1
=
ẋ2 −2 −3 x2

Solution

−1
expAt = L−1 (sI − A)

" #
−1 1 s+3 1
(sI − A) = 3
s + 3s + 2 −2 s
" #
1 s+3 1
=
(s + 2) (s + 1) −2 s
" # " #
s+3 1 2 1 1 1
(s+2)(s+1) (s+2)(s+1) (s+1) − s+2 (s+1) − s+2
= −2 s
= −2 2 −1 2
(s+2)(s+1) (s+2)(s+1) (s+1) + s+2 (s+1) − s+2

" #
2 1 1 1
At −1 −1 −1 (s+1) − s+2 (s+1) − s+2
exp =L (sI − A) =L −2 2 −1 2
(s+1) + s+2 (s+1) − s+2

" #
2 exp−t − exp−2t exp−t − exp−2t
=
−2 exp−t +2 exp−2t − exp−t +2 exp−2t

Non Homogeneous state equations

Consider the non-homogeneous state equation

ẋ (t) = ax (t) + Bu (t) (8)

Multiplying equation (52) by exp−at both sides

exp−at ẋ (t) = exp−at ax (t) + exp−at Bu (t)

exp−at [ẋ (t) − ax (t)] = exp−at Bu (t)

d 
exp−at x (t) = exp−at Bu (t)

(9)
dt
Integrating equation (53) between0 and t results in

24
ˆ t
exp−at x (t) = exp−at Bu (t) dt + x (0)
0

ˆ t
at
x (t) = exp x (0) + exp at
exp−aτ Bu (τ ) dτ (10)
0

From equation (54)

• The first term on the right hand side is the response to the initial conditions

• The second term is the response to the input u (t)

Extending the same approach to the solution of homogeneous state equation yields

ˆ t
x (t) = expAt x (0) + expA(t−τ ) Bu (τ ) dτ (11)
0

The solution of x (t) is the sum of a term consisting of the transition of the initial state and a term
arising from the input vector
Example 1.13
Obtain the time response of the following system
" # " #" # " #
ẋ1 0 1 x1 0
= + u
ẋ2 −2 −3 x2 1

where u (t) is a unit step input


Solution

ˆ t
At
x (t) = exp x (0) + expA(t−τ ) Bu (τ ) dτ
0

From the previous example


" #
At 2 exp−t − exp−2t exp−t −2 exp−2t
exp =
−2 exp−t +2 exp−2t − exp−t +2 exp−2t
" #
2 exp−(t−τ ) − exp−2(t−τ ) exp−(t−τ ) −2 exp−2(t−τ )
expA(t−τ ) =
−2 exp−(t−τ ) +2 exp−2(t−τ ) − exp−(t−τ ) +2 exp−2(t−τ )
" #
0 h i
expA(t−τ ) Bu (τ ) = expA(t−τ ) 1
1
" #
exp−(t−τ ) −2 exp−2(t−τ )
=
− exp−(t−τ ) +2 exp−2(t−τ )

25
ˆ t ˆ t
exp−(t−τ ) −2 exp−2(t−τ )
expA(t−τ ) Bu (τ ) dτ = dτ
0 0 − exp−(t−τ ) +2 exp−2(t−τ )
" #
1
2 − exp−t + 12 exp−2t
=
exp−t − exp−2t
" #
At
1
2 − exp−t + 21 exp−2t
x (t) = exp x (0) +
exp−t − exp−2t
" #" # " #
2 exp−t − exp−2t exp−t −2 exp−2t x1 (0) 1
2 − exp−t + 12 exp−2t
= +
−2 exp−t +2 exp−2t − exp−t +2 exp−2t x2 (0) exp−t − exp−2t

9 Controllability and Observability

These tells us whether it is at all possible to control all the states of the system completely by
suitable choice of input and whether it is possible to reconstruct the states of a system from its
input and outputs

Controllability

For the linear time invariant system

ẋ (t) = Ax (t) + Bu (t)

y (t) = Cx (t) (1)

The system is said to be controllable if it is possible to find some input u (t) that will transfer
the initial state of the system x (0) to the origin of the state space, x (t0 ) = 0 with t0 finite. The
solution of the state equation yields

ˆ t
x (t) = Φ (t) x (0) + Φ (t − τ ) Bu (τ ) dτ (2)
0

where Φ (t) = expAt


For the system to be controllable

ˆ t
x (0) = Φ (t0 ) x (0) + Φ (t0 − τ ) Bu (τ ) dτ = 0 (3)
0

with finite t0
A linear time invariant continuous time system is completely controllable iff the RANK of the
controllability matrix M is equal ton

26
h i
M= B AB A2 B ... An−1 B (4)

The rank of a matrix A is the maximum number of linearly independent columns of A; that is, it
is the order of the largest non singular matrix contained in A. This implies that the controllability
matrix M must be non singular for the system to be completely controllable.
If a system is not completely controllable, it implies that it has one or more natural modes that
cannot be affected by the input directly or indirectly.
Example 1.14
Determine whether the system represented by the given state space is controllable
" # " #" # " #
ẋ1 0.5 0 x1 0
= + u (t)
ẋ2 0 −2 x2 1

Solution
The controllability matrix is given by

h i
M= A AB

" #" # " #


0.5 0 0 0
AB = =
0 −2 1 −2
" #
0 0
M=
1 −2

0 0
M = =0

1 −2

The matrix is singular hence the system is uncontrollable


This is more obvious if we write the two differential equations separately as

ẋ1 = 0.5x1

ẋ2 = −2x2 + u (t)

It is evident that whereas x2 can be changed byu (t) the state x1 is unaffected by our choice of the
inputs since it is not coupled either directly to the input or to the state x2 hence this state of
x1 (0) exp−0.5t is uncontrollable
On the other hand if we had

27
ẋ1 = 0.5x1 + x2

ẋ2 = −2x2 + u (t)

The controllability matrix is obtained as


" #
0 1
M=
1 −2

0 1
M = = −1

1 −2

The matrix is nonsingular hence the system is controllable.


x1 can be controlled indirectly through x2

Observability

The linear time invariant system is said to be observable if the initial condition x (0)can be deter-
mined from the output function y (t) for 0 < t < t1 where t1 is finite

y (t) = Cx (t)

ˆ t
= CΦ (t) x (0) + C Φ (t − τ ) Bu (τ ) dτ (5)
0

Thus given u (t) and y (t)for 0 < t < t1 with t1 being some finite value, the system is observable if
equation (60) can be solved forx (0)
The system is observable if the observability matrix N is nonsingular i.e. the rank of N is equal
ton


C

CA

N = .

..

CAn−1

Example 1.15
Consider the system represented by
" # " #" # " #
ẋ1 0.5 0 x1 0
= + u (t)
ẋ2 0 −2 x2 1

28
" #
h i x
1
y (t) = 0 1
x2

Determine whether the system is observable


Solution
The observability matrix is given by
" #
C
N=
CA
" # " #
h i 0.5 0 0 1
CA = 0 1 =
0 −2 0 −2

0 1
N = =0

0 −2

The matrix is singular and therefore the system is unobservable


The state x1 does not affect the output nor does it affect the state x2 which is coupled to the output

10 Pole Placement Design Technique

State space design enables the design of a system having the desired closed loop poles or desired
characteristic equation
It also enables inclusion of initial conditions if necessary
Pole placement design is based on the state model of the system. We assume that all the state
variables are measurable and are available for feedback
State model equations

ẋ (t) = Ax (t) + Bu (t) (1)

The plant input u (t) is made a function of the states of the form

u (t) = f (x (t)) (2)

Equation (62) is called the control rule or control law. In pole placement design, the control law is
specified as a linear function of the states of the form

29
u (t) = −kx (t) (3)

This control law allows the poles of the closed system to be placed in any desirable location and is
expressed as

u (t) = −k1 x1 (t) − k2 x2 (t) − . . . − kn xn (t) (4)

The design problem is the specification of the desired root locations of the systems characteristic
equations and the calculations of the gains ki to yield these desired root locations.
A necessary and sufficient condition that the closed-loop poles can be placed at any arbitrary
location in the s-plane is that the system must be completely state controllable.

Determination of the MatrixK

1. Using Direct substitution method


If the system is of low order, direct substitution of matrixK into the desired characteristic
polynomial may be simpler.
e.g. if n = 3 and the desired poles are µ1 , µ2 and µ3 then

h i
K = k1 k2 k3 (5)

Desired characteristic polynomial

(s − µ1 ) (s − µ2 ) (s − µ3 )

This is also obtained as

|[sI − A + BK]|

We equate

(s − µ1 ) (s − µ2 ) (s − µ3 ) = |[sI − A + BK]|

to obtain the values of ki

30
Example 1.16
Consider the system

ẋ (t) = Ax (t) + Bu (t)

where  
0 1 0
A= 0 0 1
 

−1 −5 −6
 
0
B = 0
 

The system uses the state feedback control law u (t) = −kx (t). It is desired to have closed
loop poles at s = −2 ± j4 , s = −10.
Determine the state feedback gain matrix K
Solution
We first check for controllability of the system

h i
M= B AB A2 B


0 0 1

M = 0 1 −6 = −1


1 −6 31

The matrix is non singular hence the system is completely state controllable
Next we solve forK
Let h i
K = k1 k2 k3

|[sI − A + BK]|

    
s 0 0 0 1 0 0 h
i
= 0 s 0 −  0 0 1  + 0 k1 k2 k3
     

0 0 s −1 −5 −6 1

= s3 + (6 + k3 ) s2 + (5 + k2 ) s + (1 + k1 ) .........(i)

31
The desired characteristic equation is

(s + 2 − j4) (s + 2 + j4) (s + 10)

= s3 + 14s2 + 60s + 200.........(ii)

Comparing equations (i) and (ii)

k3 = 8 k2 = 55 k1 = 199

h i
K = 199 55 8

u = −199x1 − 55x2 − 8x3

2. Using Ackermann’s Formula


Cayley Hamilton Theorem
Every square matrix satisfies its own characteristic equation i.e. if the characteristic equation
of the nth order square matrix is

|λI − A| = 0

|λI − A| = λn + α1 λn−1 + α2 λn−2 + · · · + αn−1 λ + αn I = 0 (6)

Then

An + α1 An−1 + α2 An−2 + · · · + αn−1 A + αn I = 0 (7)

To obtain the inverse of the matrix A we divide equation byA on both sides

An−1 + α1 An−2 + α2 An−3 + · · · + αn−1 I + αn A−1 = 0

−1  n−1
A−1 = + α1 An−2 + α2 An−3 + · · · + αn−1 I

A (8)
αn

Example 1.17
Determine the inverse of the matrix

32
 
1 2 7
A = 4 2 3
 

1 2 1
using Cayley Hamilton Theorem
Solution

|λI − A| = 0


λ − 1 −2 −7

−4 λ−2 −3 = 0


−1 −2 λ − 1

λ3 − 4λ2 − 20λ − 35 = 0

−1  2
A−1 =

A − 4A − 20I
35

 
−4 11 −5
−1 
= −1 −6 25 

35
6 1 −10

Consider the state equation


ẋ = Ax + Bu (9)

Where the state feedback control law

u (t) = −kx (t)

We assume that the system is completely state controllable and that the desired closed loop
poles are at
s = µ1, s = µ2 . . . . . . s = µn

Equation(69) becomes

ẋ = [A − Bk] x (10)

Defining  = A − Bk
Then the desired characteristic equation is

sI − Â = 0

33
(s − µ1 ) (s − µ2 ) . . . . . . (s − µn ) = 0

sn + α1 sn−1 + α2 sn−2 + · · · + αn−1 s + αn = 0

Use of Cayley Hamilton theorem which states that  satisfies its own characteristic equation
we obtain

 
Φ Â = Ân + α1 Ân−1 + α2 Ân−2 + · · · + αn−1 Â + αn I = 0 (11)

To derive Ackermann’s formula, we consider the case when n = 3


Considering that
I=I (12)

 = A − Bk (13)

2
Â2 = [A − Bk]

= A2 − 2ABK − B 2 K 2

= A2 − ABK − BK Â (14)

2
Â3 = [A − Bk] [A − BK]

h i
= A2 − ABK − BK Â [A − BK]

= A3 − A2 BK − ABK Â − BK Â2 (15)

Multiplying both sides of equations (72), (73), (74) and (75)with, α3 ,α2 α1 and α0 where
α0 = 0

α3 I = α3 I

α2 Â = α2 [A − Bk]

34
h i
α1 Â2 = α1 A2 − ABK − BK Â

Â3 = A3 − A2 BK − ABK Â − BK Â2

Adding the two sides of the equations

Â3 +α1 Â2 +α2 Â+α3 I = A3 +α1 A2 +α2 A+α3 I−α2 BK−α1 ABK−α1 BK Â−A2 BK−ABK Â−BK Â2
(16)
Since

 
Â3 + α1 Â2 + α3 I = Φ Â = 0

And

A3 + α1 A2 + α2 A + α3 I = Φ (A) 6= 0

then

Φ (A) − α2 BK − α1 ABK − α1 BK Â − A2 BK − ABK Â − BK Â2 = 0 (17)

h i h i
Φ (A) = B α2 K − α1 K Â − K Â2 + AB α1 K − K Â + A2 BK

 2

h i α2 K − α1 K Â − K Â
Φ (A) = B AB A2 B  α1 K − K Â (18)
 

K
 
h i−1 α2 K − α1 K Â − K Â2
B AB A2 B Φ (A) =  α1 K − K Â (19)
 

K
h i
Premultiplying both sides of (79) by 0 0 1 and rearranging we obtain

h ih i−1
K= 0 0 1 B AB A2 B Φ (A) (20)

For an arbitrary positive integer n, then

h ih i−1
K= 0 0 ... 1 B AB ... An−1 B Φ (A) (21)

35
Example 1.18
Consider the system

ẋ (t) = Ax (t) + Bu (t)

where  
0 1 0
A= 0 0 1
 

−1 −5 −6
 
0
B = 0
 

The system uses the state feedback control law u (t) = −kx (t). It is desired to have closed
loop poles at
s = −2 ± j4 , s = −10.
Determine the state feedback gain matrix K using Ackermann’s formular
Solution
The desired characteristic equation is given by

(s + 2 − j4) (s + 2 + j4) (s + 10) = s3 + 14s2 + 60s + 200

Φ (A) = A3 + 14A2 + 60A + 200I

where

 
0 1 0
A= 0 0 1
 

−1 −5 −6
 
199 55 8
Φ (A) =  −8 159 7 
 

−7 −43 117
 
h 0 i 0 1
2
B AB A B = 0 1 −6
 

1 −6 31

h ih i−1
K= 0 0 1 B AB A2 B Φ (A)

36
  
h i 0 0 1 199 55 8
K= 0 0 1 0 1 −6  −8 159 7 
  

1 −6 31 −7 −43 117

h i
K = 199 55 8

3. Using Transformation Matrix T


Suppose that the system defined by

ẋ = Ax + Bu

and the control law is given by u (t) = −kx (t)


The matrix K can be obtained as follows

(a) Check the controllability condition for the system. If the system is completely control-
lable then
(b) Determine the characteristic polynomial of the matrix A

|sI − A| = sn + a1 sn−1 + a2 sn−2 + · · · + an−1 s + an

Determine the values ofa1 , a2 . . . an


(c) Determine the transformation matrix T that can transform the system state equation in
controllable canonical form. If the system is already in controllable canonical form then
T =I
Else
T = MW

Where M is the controllability matrix and


 
an−1 an−2 . . . a1 1
an−2 an−3 . . . 1 0
 
 
 . .    . .
 
W = . .    . .
 
 
 . .    . .
 
 a1 1 . . . 0 0
 

1 −an−1 . . . 0 0

(d) Write the desired characteristic polynomial

(s − µ1 ) (s − µ2 ) . . . . . . (s − µn ) = sn + α1 sn−1 + α2 sn−2 + · · · + αn−1 s + αn

37
Determine the values of α1 , α2 . . . αn
(e) Obtain the matrix K as
h i
K = αn − an αn−1 − an−1 α1 − a1 T −1 (22)

Example 1.19
Consider the system

ẋ (t) = Ax (t) + Bu (t)

where  
0 1 0
A= 0 0 1
 

−1 −5 −6
 
0
B = 0
 

The system uses the state feedback control law u (t) = −kx (t). It is desired to have closed
loop poles at
s = −2 ± j4 , s = −10.
Determine the state feedback gain matrix K using transformation matrix T
Solution
Check the system for controllability

h i
M= B AB A2 B


0 0 1

M = 0 1 −6 = −1


1 −6 31

The matrix is nonsingular hence the system is completely state controllable


Obtain the characteristic polynomial of the matrix A

|[sI − A]|

   
s 0 0 0 1 0

= 0 s 0 −  0 0 1 
   

0 0 s −1 −5 −6

38
= s3 + 6s2 + 5s + 1

Comparing with

s3 + a1 s2 + a2 s + a3

then

a1 = 6, a2 = 5, a3 = 1

The desired characteristic polynomial is given by

(s + 2 − j4) (s + 2 + j4) (s + 10) = s3 + 14s2 + 60s + 200

Comparing this with s3 + α1 s2 + α2 s + α3 then

α1 = 14, α2 = 60, α3 = 200

Since the system is already in controllable canonical form then T = I


The matrix Kis obtained as

h i
K = α3 − a3 α2 − a2 α1 − a1 T −1

h i
K = 199 55 8

11 State Observer Design

In the pole placement design approach, we assumed that all state variables are available for feedback.
In practice, however, not all the state variables are available for feedback. Then we need to estimate
unavailable state variables. A state observer estimates the state variables based on the measurement
of output and control variables
If a state observer estimates all state variables of the system, regardless of whether some state
variables are available for direct measurement, it is called a full order state observer. A necessary
and sufficient condition for observer design is that the system must be completely state observable.
Consider the system
ẋ = Ax + Bu

y = Cx (1)

39
The state variables can be estimated from the measured output and control variables

where
G-state observer gain matrix
ỹ-estimated output
x̃-estimated state variables
For the observer

˜ = Ax̃ + G (y − ỹ) + Bu

ỹ = C x̃ (2)

But y = Cx and ỹ = C x̃
Equation (84) becomes

˜ = Ax̃ + GC (x − x̃) + Bu
ẋ (3)

Subtracting (85) from (83) we obtain

˜ = A [x − x̃] − GC [x − x̃]
ẋ − ẋ

y − ỹ = C [x − x̃] (4)

˜ = ẋand
Taking x − x̃ = x̂ , ẋ − ẋ ˆ y − ỹ = ŷequation (86) becomes

ˆ = [A − GC] x̂

ŷ = C x̂ (5)

40
We can choose appropriate Eigen values of [A − GC] to enable placement of poles of the closed
loop system at desired locations
The control design problem is to determine the matrix G;(n ∗ 1) matrix
where

 
g1
 g2 
 
G=
 .. 

.
gn

Determining the matrix G

1. Using Direct Substitution method


Similar to the case of pole placement, if the system is of low order, then direct substitution
of the matrix G into the desired characteristic polynomial may be simpler. e.g. if x is a 3
vector then G can be written as  
g1
G = g2 
 

g3

Substituting this G into the desired characteristic polynomial

|sI − (A − GC)| = (s − µ1 ) (s − µ2 ) (s − µ3 )

By equating the coefficients of the powers on both sides of this equation, we can obtain the values
of g1 , g2 and g3
Example 1.20
Consider the system
ẋ = Ax + Bu

y = Cx

Where
" #
0 20.6
A=
1 0
" #
0
B=
1

h i
C= 0 1

41
Design a full order state observer assuming that the desired Eigen values of the observer matrix
are µ1 = −10 µ2 = −10
Solution
Test the system for observability
Observability matrix " # " #
C 0 1
N= =
CA 1 0

|N | = −1

The system is completely state observable and determination of observer gain matrix is possible
Let " #
g1
G=
g2

|sI − (A − GC)| = 0


s −20.6 + g1
= s2 + g2 s − 20.6 + g1 = 0


−1 s + g2

The desired characteristic equation is given by

(s − µ1 ) (s − µ2 ) = (s + 10) (s + 10) = 0

s2 + 20s + 100 = 0

Comparing the two equations


g2 = 20and g1 = 120.6 " #
120.6
G=
20

Using Ackermann’s Formula


The observer gain matrix is given by

 −1  
0C
 CA  0
   
G = Φ (A) 
 ..   .. 
  
 .  .
CAn−1 1

Where

42
Φ (A) = An + α1 An−1 + α2 An−2 + · · · + αn−1 A + αn I

Example 1.21
Determine the observer gain matrix for example 1.20 using Ackermann’s formula
Solution

" #−1 " #


C 0
G = Φ (A)
CA 1
" #
0 20.6
A=
1 0

h i
C= 0 1

The desired characteristic equation is given by

(s − µ1 ) (s − µ2 ) = (s + 10) (s + 10)

s2 + 20s + 100

Φ (A) = A2 + 20A + 100I

" #2 " # " #


0 20.6 0 20.6 1 0
= + 20 + 100
1 0 1 0 0 1
" #
120.6 412
=
20 120.6
" #" #" # " #
120.6 412 1 0 0 120.6
G= =
20 120.6 0 1 1 20

Using Transformation Matrix Q


Following the same procedure as in deriving the state feedback matrix K then

1. Check the observability condition for the system. If the system is completely observability
then

2. Determine the characteristic polynomial of the matrix A

|sI − A| = sn + a1 sn−1 + a2 sn−2 + · · · + an−1 s + an

43
Determine the values ofa1 , a2 . . . an

3. Determine the transformation matrix Q that can transform the system state equation in
controllable canonical form. If the system is already in observable canonical form then Q=I
Else
Q = WN

WhereN is the observability matrix and

 
an−1 an−2 . . . a1 1
an−2 an−3 . . . 1 0
 
 
 . .    . .
 
W = . .    . .
 
 
 . .    . .
 
 a1 1 . . . 0 0
 

1 −an−1 . . . 0 0

4. Write the desired characteristic polynomial

(s − µ1 ) (s − µ2 ) . . . . . . (s − µn ) = sn + α1 sn−1 + α2 sn−2 + · · · + αn−1 s + αn

Determine the values of α1 , α2 . . . αn

5. Obtain the observer gain matrix G as


 
αn − an
αn−1 − an−1 
 
G = Q
 .. 
 (6)
 . 
α1 − a1

Example 1.21
Determine the observer gain matrix for example 1.20 using the transformation matrix method
Solution
Observability has already been tested
" #
0 20.6
A=
1 0

h i
C= 0 1

" #
s −20.6
|sI − A| = = s2 − 20.6
−1 s

44
Comparing this with s2 + a1 s + a2
a1 = 0, a2 = −20.6
The system is already in state observable form, therefore
" #
1 0
Q=I=
0 1

The desired characteristic equation is

(s − µ1 ) (s − µ2 ) = (s + 10) (s + 10) = 0

s2 + 20s + 100 = 0

From this we obtain the values of α1 = 20 , α2 = 100


" #
α2 − a2
G=Q
α1 − a1
" #" # " #
1 0 120.6 120.6
= =
0 1 20 20

45
2 DISCRETE SYSTEMS

1 Sampled Data Systems

A sampled data system operates on discrete-time rather than continuous-time signals. A digital
computer is used as the controller in such a system. A D/A converter is usually connected to the
output of the computer to drive the plant.

The Sampling Process

A sampler is basically a switch that closes every T seconds as shown below

When a continuous signal r(t) is sampled at regular intervals T, the resulting discrete signal is as
shown in Figure where q is the amount of time that the switch is closed.

In control applications, the switch closure time q is much smaller than the sampling time T and
can be neglected.
The ideal sampling process can be considered as the multiplication of a pulse train with a continuous
signal

r∗ (t) = p(t)r(t)

46
where p(t) is the delta pulse train

thus
X
p(t) = δ(t − nT )
X
r∗ (t) = r(t) δ(t − nT )

It is possible to reconstruct r(t) from r∗ (t) by use of a hold device, the most common of which is
the zero order hold.
This circuit remembers the last information until a new sample is obtained i.e. zero order hold
takes the value r(nT ) and holds it constant for nT ≤ t ≤ (n + 1)T and the value of r(t) is used
during the sampling period.

It can be seen that the zero order hold converts a series of impulses into a series of pulses of width
T. Hence a unit impulse at time t is converted into a pulse of width T, which may be created by a
positive unit step at time t, followed by a negative unit step at time t-T i.e. delayed by T.
The transfer function of a zero order hold is

1 1 −T s
L{r(t)} = − e
s s

1 − e−T s
=
s

47
2 The Z-Transform

Z-transform is the principal analytical tool for SISO discrete time systems, and is analogous to the
Laplace transform for continuous time systems. The symbol Z can be associated with discrete time
shifting in a difference equation in the same way in a differential equation
Given that

X
r∗ (t) = r(kT )δ(t − kT )
k=−∞

Taking the Laplace transform of this equation which is the ideal sampled signal gives

X
R∗ (s) = L{r∗ (t)} = r(kT )e−kT s
k=0

Defining z as z = esT

X
R∗ (z) = r(kT )z −k
k=0

R(z) = Z{r(t)}

Notice that R(z) consists of an infinite series in the complex variable z and

R(z) = r(0) + r(T )z −1 + r(2T )z −2 + · · ·

i.e. r(kT ) are the coefficients of this power series at different sampling instants
Example
Unit step

 0 k<0
r(kT ) =
 1 k≥0


X
R(z) = r(kT )z −k
k=0


X
= (1)z −k
k=0

= 1 + z −1 + z −2 + z −3 + · · ·

z
=
z−1
Unit Ramp

48

 0 k<0
r(kT ) =
 kT k≥0


X
R(z) = r(kT )z −k
k=0


X
= (kT )z −k
k=0

= T z −1 + 2T z −2 + 3T z −3 + · · ·

Tz
= 2
(z − 1)

Exponential function

 0 k<0
r(kT ) =
 e−akT k≥0


X
R(z) = r(kT )z −k
k=0


X
= (e−akT )z −k
k=0

= 1 + e−aT z −1 + e−2aT z −2 + e−3aT z −3 + · · ·

z
=
z − e−aT
Sine function

 0 n<0
r(kT ) =
 sin nωT n≥0

ejnωT − e−jnωT
r(kT ) =
2j

ejnωT e−jnωT
= −
2j 2j

49
z
R(e−anT ) =
z − e−aT
 
1 z z
R(z) = −
2j z−e jωT z − e−jωT

z sinωT
=
z2 − 2z cosωT + 1
Cosine function

 0 n<0
r(kT ) =
 cos nωT n≥0

ejnωT + e−jnωT
r(kT ) =
2
ejnωT e−jnωT
= +
2 2

z
R(e−anT ) = R (z) =
z − e−aT
 
1 z z
R(z) = +
2 z−e jωT z − e−jωT
 
z (z − cos ωT )
=
(z − ejωT ) (z − e−jωT )

z (z − cos ωT )
=
z2 − 2z cosωT + 1
Discrete impulse function

1 n=0
δ (n) =
0 n 6= 0


X
R (z) = r (nT ) z −n
n=0


X
= (1) z −n
n=0

=1

Delayed discrete impulse function

50

1 n=k>0
δ (n − k) =
0 n 6= k


X
R (z) = r (nT ) z −n
n=0


X
= (1) z −n
n=0

= z −n

commonly used Z transform


(Table of z transform)
Example
Determine the Z transform given that

1
G(s) =
s2 + 5s + 6
Method 1: Using Inverse Laplace Transform

1 1 1
G(s) = = −
s2 + 5s + 6 s+2 s+3

L−1 {G(s)} = e−2t − e−3t



X
G(z) = (e−2nT − e−3nT )z −k
k=0

z z
= −
z − e−2T z − e−3T
Method 2: Using Z transform table
from z transform table

1 z
=
s+a z − e−aT
z z
G (z) = −
z − e−2T z − e−3T

example 2

1
X (s) =
s (s + 1)

51
solution
A B As + A + Bs
X (s) = + =
s s+1 s (s + 1)
A=1 B = −1

1 1
X (s) = −
s s+1

x (t) = 1 − e−t

X (kT ) = 1 − e−kT


X
1 − e−kT z −k
 
X (z) =
k=0


X ∞
X
= (1) z −k − e−kT z −k
k=0 k=0

1 1
= −
1 − z −1 1 − e−T z −1

z z

z − 1 z − e−T

z 1 − e−T
 
=
(z − 1) (z − e−T )

3 Inverse Z Transform

The inverse Z transform of X (z) yields the corresponding time sequence x (t)
The notation for the inverse Z transform is z −1

Direct division method

We obtain the inverse Z-transform by expanding X (z) into an infinite power series in z −1


X
X (z) = x (kT ) z −k
k=0

= x (0) + x (T ) z −1 + x (2T ) z −2 + · · ·

The values of the coefficient of z −1 can then be obtained by inspection

52
Example
Determine the values of x (k) for x (z) given by

z
x (z) =
z 2 − 3z + 2

solution
z −1 + 3z −2 + 7z −3 + 15z −4
X (z) = z 2 − 3z + 2 z
z − 3 + 2z −1
3 − 2z −1
3 − 9z −1 + 6z −2
7z −1 − 6z −2
7z −1 − 21z −2 + 14z −3
15z −2 − 14z −3
15z −2 − 45z −3 + 30z −4
31z −3 − 30z −4

X (z) = z −1 + 3z −2 + 7z −3 + 15z −4 + · · ·

x (0) = 0 x (1) = 3 x (2) = 7 x (3) = 15

Example 2
A sampled data system has the transfer function

1
G (s) =
s+1

If the sampling time is one second and the system is subject to one unit step input function ,
determine the discrete time response
Solution
X (z) = G (z) Xi (z)
z z
= −T
z−e z−1
z z
=
z − 0.368 z − 1

z2
=
z 2 − 1.368z + 0.368

53
1 + 1.368z −1 + 1.503z −2
2
z − 1.368z + 0.368 z2
z 2 − 1.368z + 0.368
1.368z − 0.368
1.368z − 1.871 + 0.503z −1
1.503 − 0.503z −1
1.503z −2 − 2.056z −1 + 0.553z −2

= 1 + 1.368z −1 + 1.503z −2 + · · ·

x (0) = 1 x (1) = 1.368 x (2) = 1.503

Partial fraction expansion method

The function X (z) may be expanded into partial fractions in the same manner used in Laplace
transform
Example
Obtain the inverse Z transform of X (z) given that

z
X (z) =
z 2 − 3z + 2

and sampling time T = 1


Solution

z z
X (z) = =
z2 − 3z + 2 (z − 1) (z − 2)

X (z) 1 A B
= 2 = +
z z − 3z + 2 z−1 z−2

−1 1
= +
z−8 z−2

−z z
X (z) = +
z−1 z−2
 
−z z
z −1 [X (z)] = z −1 +
z−1 z−2
from z transform tables  
z
z −1 =1
z−1

54
 
z
z −1 = e0.69
z−2

x (k) = e0.69 − 1

= 2k − 1

Example 2
Given that
1 − e−aT z

X (z) =
(z − 1) (z − eaT )
where a is a constant and T the sampling time , determine the inverse z transform
Solution

1 − e−aT

X (z)
=
z (z − 1) (z − eaT )

A B
= +
(z − 1) (z − eaT )

1 1
= −
(z − 1) (z − eaT )

z z
X (z) = −
(z − 1) (z − eaT )
   
z z
z −1 [X (z)] = z −1 − z −1
(z − 1) (z − eaT )

= 1 − eakT

x (kT ) = 1 − eakT k = 0, 1, 2, ....

Pulse Transfer Function

Pulse transfer function relates the Z transform of the output at the sampling instants to that of
the sampled input

55
X0 (z)
G (z) =
U (z)
The ratio of the output X0 (z) and input U (z) is called the pulse transfer function of the discrete
time system

Blocks in cascade

There are synchronized samplers either side of the blocks G1 (s) and G2 (s) . The pulse transfer
function is therefore

Y
(z) = G1 (z) G2 (z)
U

There is no sampler between G1 (s) and G2 (s) and so they can be combined to give G1 G2 (s)
The output is given by
Y (z) = z [G1 G2 (s)] U (z)

PTF
Y (z)
= z [G1 G2 (s)]
U (z)

= G1 G2 (z)

Example
Consider the systems shown where

1 1
G1 (s) = G2 (s) =
s+a s+b

Obtain the pulse transfer function for each of the two systems

56
(a) We assume that the two samplers are synchronized and have the same sampling period
The PTF is given by
Y (z)
= G1 (z) G2 (z)
U (z)
   
1 1
=z z
s+a s+b

z z
= .
z − e−aT z − e−bT

z2
=
(z − e−aT ) (z − e−bT )
(b)
Y (z)
= G1 G2 (z)
U (z)
 
1 1
=z .
s+a s+b

1
=
(s + a) (s + b)
A B 1
+ =
s+a s+b (s + a) (s + b)

B (s + a) + A (s + b) = 1

1 −1
s = −a A= =
−a + b a−b

1 1
s = −b B= =
−b + a a−b
  
1 −1 1
⇒z +
a−b s+a s+b
 
1 −z z
= +
a − b z − e−aT z − e−bT
" #
−z z − e−bT + z z − e−aT

1
=
a−b (z − eaT ) (z − e−bT )
"  #
1 z e−bt − e−at
=
a − b (z − eaT ) (z − e−bT )

57
Pulse transfer function of closed-loop systems

The existence or non-existence of an output sampler in the loop makes a difference in the behavior
of the system

1. with no output sampler

The closed loop pulse transfer function is given by

C (z) G (z)
=
R (z) 1 + GH (z)

where
GH (z) = z {GH (s)}

2. with output sampler

The closed loop pulse transfer is given by

C (z) G (z)
=
R (z) 1 + G (z) H (z)

Example
The figure below shows a digital control system. When the controller gain k is unity and the
sampling time is 0.5 seconds, determine

1. the open loop pulse transfer function

2. the closed loop pulse transfer function

58
3. the difference equation for the discrete time response

Solution
(a)
1 − e−T s
  
1
G (s) = k
s s (s + 2)
but k = 1  
1
= 1 − e−T s

s2 (s + 2)
partial fraction expansion
1 A B C
= + 2+
s2 (s + 2) s s s+2

A = −0.25 B = 0.5 C = 0.25

 
−0.25 0.5 0.25 
1 − e−T s

G (s) = + 2 +
s s s+2
 
 −T s
 −1 2 1
= 0.25 1 − e + 2+
s s s+2

" #
−z 2T z z
G (z) = 0.25 1 − z −1
 
+ +
z − 1 (z − 1)2 z − e−2T

but T = 0.5  " #


z−1 −z 2T z z
G (z) = 0.25 + +
z z − 1 (z − 1)2 z − 0.368

0.092z + 0.066
=
z 2 − 1.368z + 0.368
(b) closed loop pulse transfer function

C G (z)
(z) =
R 1 + G (z)

0.092z + 0.066
=
z2 − 1.368z + 0.368 + 0.092z + 0.066

59
0.092z + 0.066
=
z2 − 1.276z + 0.0434

C (z) 0.092z −1 + 0.066z −2


=
R (z) 1 − 1.276z −1 + 0.434z −2

C (z) 1 − 1.276z −1 + 0.434z −2 = R (z) 0.092z −1 + 0.066z −2


   

C (kT ) − 1.276C (k − 1) T + 0.434C (k − 2) T = 0.092r (k − 1) T + 0.066r (k − 2) T

C (kT ) = 0.092r (k − 1) T + 0.066r (k − 2) T + 1.276C (k − 1) T − 0.434C (k − 2) T

Stability in the z-plane

Just as the transient analysis of continuous systems may be undertaken in the s-plane , stability
and transient analysis on discrete systems may be conducted in the z-plane
It is possible to map from the s-plane to z-plane using the relationship

z = esT

s = σ ± jω

⇒ z = e(σ±jω)T

= eσT ejωT

= eσT ej(ωT +2πk)

This means that there are infinitely many values of s for each value of z
For stability in s-plane σ must be on the left half of the s-plane ie σis negative. Thus

|z| = eσT < 1



If T = ωS where ωs is the sampling frequency

j 2πω
z = |z| e ωS

60
It can be seen that the left hand side (stable) of the s-plane corresponds to a region within a circle
of unity radius in the z plane

The necessary and sufficient condition for the stability of a DTs is that all the poles of its z transform
function lie inside the unit circle of the z plane
Example
Consider the closed loop system whose open loop transfer function is given by

1 − e−sT 1
G (s) =
s s (s + 2)

Determine the stability of the system where k = 1 , T = 0.5

1 − e−sT 1
G (s) =
s s (s + 2)

0.092z + 0.066
G (z) =
z2 − 1.368z + 0.368

C (z) G (z)
=
R (z) 1 + G (z)

61
0.092z + 0.157
=
z2 − 1.276z + 0.527

C.E = 1 + G (z) = z 2 − 1.276z + 0.527

z = 0.63 ± j0.34

= 0.72

|z| < 1 stable

Routh Hurwitz Criterion

It is possible to use Routh Hurwitz criterion to determine whether a polynomial ϕ (z) has roots
outside the unit circle.
We use the transformation
ω = u + jv

z+1
=
z−1

ω+1
z=
ω−1
The characteristic polynomial ϕ (z) is transformed into a polynomial of ω and the Routh Hurwitz
criterion applied to determine the stability of the system
Example
Determine the stability of the system above using Routh Hurwitz criterion

Jury’s Stability test

Let the characteristic equation of a discrete system be

ϕ (z) = an z n + an−1 z n−1 + · · · + a1 z + a0 = 0

where an > 0
The Jury’s array is formed as

62
z0 z1 z2 ··· z n−k ··· z n−1 zn
a0 a1 a2 ··· an−k ··· an−1 an
an an−1 an−2 ··· ak ··· a1 a0
b0 b1 b2 ··· bn−k bn−1
bn−1 bn−2 bn−3 ··· bk−1 b0
c0 c1 c2 ··· cn−k
cn−2 cn−3 cn−4 ··· ck−2
.. .. .. ..
. . . .
The elements of each of the even numbered rows are the elements of the preceding row in reverse
order
The elements of the odd numbered row are defined as

a an−k
0
bk =
an ak

b bn−1−k
0
ck =
bn−1 bk

c cn−2−k
0
dk =
cn−2 ck

The necessary and sufficient conditions for the polynomial ϕ (z) to have no roots outside or on the
unit circle with an > 0 are as follows

ϕ (1) > 0

n
(−1) ϕ (−1) > 0

|a0 | < an

|b0 | > |bn−1 |

|c0 | > |cn−2 |

|d0 | > |dn−3 |

..
.

63
|m0 | > |m2 |

NB
For a second order system , the array contains only one row. For each additional order, two
additional rows are added to the array
For an nth order system , there are a total of n + 1 constraints
Jury’s test may be applied as follows

n
1. Check the three conditions ϕ (1) > 0 , (−1) ϕ (−1) > 0 and |a0 | < an which requires no
calculation. Stop if any of this conditions are not satisfied

2. Construct the array, checking the other condition as each row is calculated. Stop if any
condition is not satisfied

Example
Suppose the characteristic equation of a closed loop discrete time system is given by

ϕ (z) = z 3 − 1.8z 2 + 1.05z − 0.20 = 0

solution
n=3

ϕ (1) = 1 − 1.8 + 1.05 − 0.20 = 0.05

ϕ (1) = 0.05 > 0

3
(−1) ϕ (−1) = (−1) (−1 − 1.8 − 1.05 − 0.20) = 4.05

3
(−1) ϕ (−1) > 0

|a0 | = 0.2

a3 = 1

|a0 | < a3

conditions (i) , (ii) and (iii) are met

64
Jury’s array
z0 z1 z2 z3
−0.2 1.05 −1.8 1
1 −1.8 1.05 −0.2
−0.96 1.59 −0.69

−0.2 1
b0 =


1 −0.2

|b0 | = 0.96


a
0 a2 −0.2 −1.8
b1 = = = 1.59
a3 a1 1 1.05

a
0 a1 −0.2 1.05
b2 = =
a3 a2 1 1.8

|b2 | = 0.69

|b0 | > |b2 |

All the conditions are satisfied; the system is stable


The characteristic equation can be factorized as

2
ϕ (z) = (z − 0.5) (z − 0.8)

z2 = z1 = 0.5 z3 = 0.8

z1 < 1 z2 < 1 z3 < 1

All poles lie inside the unit circle hence the system is stable
Example 2
Consider the system described with characteristic equation

z 2 − 1.276z + 0.0434 = 0

a2 = 1 a1 = −1.276 a0 = 0.043

condition

65
ϕ (1) = 1 − 1.276 + 0.0434 = −0.3194

ϕ (1) < 0

condition not met hence system unstable!!! STOP here!!!

66