State Variable Analysis

Unit 5

State variable analysis

The state-space description provide the
dynamics of a system as a
Set of coupled first-order differential
equations in a set of internal variables
known as state variables, together with
A set of algebraic equations that
combine the state variables into
physical output variables.

State variable analysis

Properties:There is no unique set of state variables that describe any
given system.
Many different sets of variables may be selected to yield
a complete system description.
State variable descriptions of systems may be formulated
in terms of physical and measurable variables, or in terms
of variables that are not directly measurable.
It is possible to mathematically transform one set of state
variables to another.

Need For State variable

analysis
Generally control system can be
Modelled, Analysed and Designed by
two ways.
1. Transfer Function approach
2. State Variable Approach.

Transfer Function approach

Transfer Function approach

Limitation:1. Applicable to LTIV System.
2. Restricted to SISO system and
Cumbersome for MIMO system.
3. Concentrated only towards only
input
and output of the system.
4. Failed to give focus on Internal
elements.

State Variable approach

Advantages:1. Applicable to LTV & LTIV System.
2. Applicable to SISO & MIMO system.
3. Improved stabilization and
performance.
4. Applicable to Nonlinear system.
5. Applicable to both continuous-time
and
Discrete-time case.

State Variable approach

Factors:1.
2.
3.
4.
5.

State
State
State
State
State

and State Variable.

equations.
Space.
Trajectory.
Model.

State Variable approach

State Variable approach

State Variable approach

State Variable
Representation

From the above relation, we get

V t

1
M

F t dt

1
M

t0

F t dt

1
F t dt

M t0

Similarly

1
V t V t0
F t dt

M t0
t

t0

t0

x t v t dt v t dt v t dt
t

1
x t x t 0 t t 0 v t0
d F t dt

M t0 t0
13

State Model Representation

In State variable formulation of the system

x1 t , x2 t , xn t

State Variables are represented by

u t , u2 t , un t

Input Variables are represented by1

y1 t , y2 t , yn t

Output Variables are represented by

u1 t
u t
2

In Vector Form

u t

, y t

y1 t
y t
2

um t

, x t

x1 t
x t
2

y p t

xn t

15

For a general system, the state variable representation can

be arranged in the form of n first order differential equations.

dx1
x1 f1 x1 , x2 , xn ; u1 , u 2 , u m
dt

dxn
xn f1 x1 , x2 , xn ; u1 , u 2 , u m
dt

Integration of above equations gives,

xi t xi t 0 f i x1 , x2 , xn ; u1 , u 2 , u m
t0

16

State Equation:The n differential equations may be written in vector form

as,

t f x t , u t

t f x t , u t , t

For Time-invariant system

xi
For Time-varying system
xi

Output Equation:The output y(t) may be written in vector form as,

t g x t , u t
For Time-varying systemy t f x t , u t , t

For Time-invariant system

y

17

State Model

From the general representation of State equation and output equati

Derivative of each state variable now becomes a linear combination
states and inputs,

In the vector matrix form,

State Model
A is nn system matrix,

B is nm INPUT matrix,

State Model

Similarly, output variables at time t are linear combination of the val

and state variables at time t,i.e.,

In the vector matrix form,

State Model
C is pn output matrix,

D is pm transmission matrix,

State Variable
Representation
Types:-

1. Phase Variable Form.

2. Canonical Form.

Phase variable Form

Cases:-

1. Transfer function has no

zero term.
2. General Transfer function.

Generally, Transfer function has the form(does not

have zeros)

Y (S )
b
T (s)

U (S )
a n s n a n 1 s n 1 a1 s a0

This corresponds to the differential equation

( n )

an y an 1

( n 1)

By Considering

y a1 y a0 y bu

It reduced to n first order

differential
equation

x1 y

x1 x2

x2 x3

x2 y

xn

( n 1)

x n1 xn

a0
a n1
a1
xn
x1
x2
x1
an
an
an

The above differential equation result in the

following state equation

x 1
x
2

x
n 1
x n

0
a1

an

0
a2

an

0
a0
an

0 x1 0

0
0
0 x2

u


0
1
0
x
an 1 n 1


an xn b

Then the output equation y = x1

y 1 0 0 0

x1
x
2

x
n1
xn

0
0

0
0
u

Generally, Transfer function has the form( both

poles
m and zeros)
m 1

b m s b n 1 s
b1 s b0
Y (S )
T ( s)

U (S )
a n s n a n 1 s n 1 a1 s a0
The Transfer function may be divided into two
parts.

Y (s)
X 1 (s) Y (s)
G (s)

U (s)
U (s) X 1 (s)

where

X 1 (s)
1

n
n 1
U (s)
a n s a n 1 s
a 1 s a0

and

Y (S )
T ( s)
b m s m b n 1 s m 1 b1 s b0
X 1 (S )

The state variable representation of the first term,

X 1 (s)
1

U (s)
a n s n a n 1 s n 1 a1 s a0
The state equation

x 1

x2

x n1
x
n
Output equation
y 1

0
0

1
0

0
1

0
a
0
an

0
a
1
an

0
a
2
an

x1
x
2

0
0
x
n 1
xn

x1 0


x
2 0

u


0
1

an1 xn1 0

an xn b
0
0

0u

The state variable representation of the Second

term,

Y (S )
T (s)
b m s m b n 1 s m 1 b1 s b0
X 1 (S )

Output equation

y b m x n b m1 x n1 b1 x 1 b0 x1
x 1 x2
x1 x3

x n1 xn
a0
a n1
a1
x n
x1
x2
x1
an
an
an

The state variable representation of the second

term,

a0
a n1
a1

y bm
x1
x2
xn bu b m1 xn1 b1 x2 b0 x1
an
an
an

a0
a n11
a1
y b0
bm x1 b1
bm x2 bm1
bm xn bu
an
an
an

Output equation

a0

y b0
bm
an

b1

a1
bm
an

a n1

bm1
an

x1

x
2

bm
b0 u

xn1
x
n

Advantages:1.

It can be obtained by inspection

from the transfer

2.

function and vice versa.

It provides powerful method of

state variable
formulation.
Limitation:1.

It is not the physical variables of

the system and

therefore these are not

Problem No-I
Represent the following differential equation in state model using
phase variable approach.

y 6 y 11 y 6 y 8u

By inspection,

a0 6 & a1 11 & a 2 6 & a3 1 & b 8

Then the overall State Model,

x 1
0

x2 0
x 3
6

y 1 0

1
11 6
1
0

x1

0 x 2 0 u
x3

x1
0

x 2 0 u
x3
8

Problem No-II
Represent the following differential equation in state model using
phase variable approach.

y 6 y 11 y 6 y u 8u
17u 8u

By inspection,

a0 6 & a1 11 & a 2 6 & a3 1

b0 8 & b1 17 & b2 8 & b3 1

Then the overall State Model,

1
x
0

x2 0
x

6
3

y 2

1
0
11

0 x1
0

1 x 2 0 u
8
6 x3

x1

2 x2 0u
x3

Problem No-III

Ri t
i (t )dt v(t )

1
d
y t i (t )dt
&
C y t i (t )
C
dt

Problem No-III

d
RC
y t y t v (t )
dt
RC
d
1
1
y t
y t
v t
dt
RC
RC
The matrix coefficient

State Model

x (t ) 1
1 RC v(t )

x1 (t )

RC

x1 (t )

1
A a0

RC
1
B b0
C 1 D 0
y (t ) 1

RC

Problem No-IV

d y (t ) R dy (t ) 1
1

y (t )
v(t )
2
dt
LC
LC
L dt

Problem No-IV

The matrix coefficient

0
A
a0

1
0

a1 1 / LC
0 0
B

b
1
/
LC

0
C 1 0 D 0

1
R / L

State Model
0
1
x 1 (t )
x (t ) 1 / LC R / L

2
x1 (t )
y (t ) 1 0

x
(
t
)
2

x1 (t ) 0
x (t ) 1 / LC v(t )

CANONICAL FORM OF STATE VARIABLE

REPRESENTATION
Advantages:1. It plays an important role in
control theory for
analysis and to get time response
of the system.
2. It represent the n first order
differential equations
are completely
independent(decouple) of each
other.
Limitation:-

Generally, Transfer function has the form( both

poles and zeros)

b m s m b n 1 s m 1 b1 s b0
Y (S )
T ( s)

U (S )
a n s n a n 1 s n 1 a1 s a0
Assume that the denominator known in factored form and the poles
of the transfer function, located at 1, 2, n. are all distinct.
n
ci
Y (s)
T (s)
bm
U (s)
i 1 s i
i i xi u , i 1,2,3......n
The state equation in canonical x
form

x 1

x2

x
0
n

0
0

1

1
1
u

n x n
1
0

x1

x
2

The output equation in canonical form

y c1 x1 c2 x2 L L cn xn bmu
In vector Matrix form.

y c1

c2

cn

x1
x
2
M
bmu

M

xn

Problem No-I

Represent the following differential equation in state model using

canonical approach.

T (s)

s 8s 17 s 8
s 6s 11s 6

By using partial fraction, The Residues and the

poles are.,

1
2
1
T ( s) 1

s 1 s 2 s 3

STATE MODEL USING CANONICAL

FORM
Y ( s)

ci
T ( s)
bm
U (s)
i 1 s i

The state equation in canonical form

x 1

x2

1
x1

1
x

1
u

x
1
n

The output equation in canonical form

y c1

c2

cn

x1
x
2
M
bmu

M

xn

Problem No-I
Then the overall State
Model,

x 1
1

0
2

x 3
0

y 1

0
2
0

0
3

x1
1

1
u
2

x3
1

x1

2 1 x2 1u
x3

Concept of Controllability and

Observability
Need:1.

To check whether a control

solution exist for a
System or not.
2. To identify the dependence of
state variable of
the system on the
input.
3. To determine the initial state of
the system by
knowing the output vector for
a finite length of
time.
4. To know whether system can be

Controllability

Observability

Testing of Controllability and

Observability
Methods of testing:1. Kalman test.[applied to any
state model]

2. Gilbert test.[applicable only

for canonical
MODEL form]

STATE

Kalman test of controllability

A general nth order multi input linear time
invariant system state equation is given by

x& Ax Bu

The system is controllable if and only

if the rank of the composite matrix is
equal to order of the system n

Qc B M AB M L

MA B
n 1

Problem No-I
Consider the system with state equation to check
system is controllable or not.

x&1
x&
2
x&3

1
0
0
0

6 11

0
1

6

0
x1
x 0 u
2
x3
1

Let us now test controllability of this system by the

Kalman test

1
0
0
0

A 0
0
1 & B
0

6 11 6
1

Problem No-I
Then

1
0
0

AB 0
0
1
6 11 6

0 0
0 1

1 6

1
0
0

2
A B 0
0
1
6 11 6

0 1
1 6

6 25

Problem No-I
Then the composite matrix

Qc B M AB M L

MA B
n 1

0 0 1

Qc 0 1 6
1 6 25

Then

det Qc 0
Its rank is r=n=3. the system is therefore
completely controllable.

Kalman test of observability

A general nth order multi input linear time
invariant system output equation is given by

y Cx

The system is observable if and only if

the rank of the composite matrix is is
equal to order of the system n

Qo C M A C M L

MA

n 1

Problem No-I
Consider the system with state equation to check
system is observable or not.

x&1
x&
2
x&3

0 1
0 0

0 2

0
1

3

x1

y 3 4 1 x2
x3

0
x1
x 0 u
2
x3
1

Problem No-I
Let us now test observability of this system by the
Kalman test

0 0 0

T T
A C 1 0 2
0 1 3

3
4

1

0 0 0
2

T
T
A C 1 0 2
0 1 3

0
1

0 0
1 2

1 2

Problem No-I

Qo C M A C M

MA

3 0 0

Qc 4 1 2
1 1 2
The rank of the matrix Q0 is r=2 while n=3.
hence one of the state variable is un-

Solution of State equation

Need:To know more about
system transient response.
Types of System[Based on the input
availability]:1. Forced system[non
homogeneous system]

Let us first recall the classical method of

solution,

dx
ax;
dt

x(0) x0

The equation has the solution:

x(t ) e at x0

a 2t 2 a 3t 3
I at

K x0
2!
3!

Solutions to the State Equations Preliminaries

Let us first recall the classical method of solution,
x&(t ) Ax(t ) Bu (t )
y (t ) Cx(t ) Du (t )
A 2t 2 A 3t 3
e I At

2!
3!
o solve these equations, we will need a few mathematical tools.
First:
At

e A ( t ) e At e A
k
e A ( t ) where
e At e At I is
I an NxN identity matrix. A is simply AxAxA.
For any real numbers t and :

d At d
A 2t 2 A 3t 3
A 3t 2
2
e I At

A A t

Further,
setting

=
-t:
dt
dt
2
!
3
!
2
!

A 2t 2 A 3t 3
At

A
I

A
t

A
e
Next:

2!
3!

x
(t ) Ax(t )
x(t ) e At x(0), t 0

Properties of state transition matrix

They are

o e

A0

t e e
At

or

At

t t

t1 t2 e

A t1 t2

At1 At2

t1 t2 t2 t1

Solutions to the State Equations Preliminaries

Recall our state equations:
x&(t ) Ax(t ) Bu (t )
y (t ) Cx(t ) Du (t )
To solve these equations, we will need a few mathematical tools.
First:
A 2t 2 A 3t 3
At
e I At

2!
3!
where I is an NxN identity matrix. Ak is simply AxAxA.
)
At A numbers t and :
For
e A ( t any
ereal
e
t At
Further,
e A ( t ) e Asetting
e I = -t:

Next:

d At d
A 2t 2 A 3t 3
A 3t 2
2
e I At

A A t

dt
dt
2!
3!
2
!

A 2t 2 A 3t 3
A I At

Ae At
2!
3!

x
(t ) Ax(t )
At
We can use these results to show that the solution
x(t ) eto
x(0), t 0
is:

Solution via the Laplace Transform

Recall our state equations:
x&(t ) Ax(t ) Bu (t )
y (t ) Cx(t ) Du (t )
Using the Laplace transform on the first equation:
sX ( s ) x(0) Ax( s) Bu ( s)

sI A X(s) x(0) Bu ( s)
1
1
X( s ) sI A x(0) sI A Bu ( s )

Comparing this to:

t

x(t ) e x(0) e A t Bu ( )d , t 0
At

reveals that:
1
e At L1 sI A
Continuing with the output equation:
y (t ) Cx (t ) Du (t )

Y ( s ) CX( s ) Du ( s )
sI A

x(0) C sI A

For zero initial conditions:

1

B D u ( s )

Y( s ) H ( s ) X( s ) where H ( s ) C sI A B D
1

The transfer
function can be
computed directly
from the system
parameters.

Solution of state equation

The solution of the state differential equation can be obtained in a
manner similar to the approach we utilize for solving a first order
differential equation. Consider the first-order differential equation

x ax bu
Where x(t) and u(t) are scalar functions of time. We expect an
exponential solution of the form eat. Taking the Laplace transform of
both sides, we have

s X(s) x 0 a X(s) b U(s)

therefore,

x (0)
b
X(s)

U(s)
sa sa
The inverse Laplace transform of X(s) results in the solution
t

x ( t ) e at x (0) e a ( t ) b u () d
0

We expect the solution of the state differential equation to be

similar to x(t) and to be of differential form. The matrix
exponential function is defined as

e At

A2t 2
Ak t k
I At

2!
k!

which converges for all finite t and any A. Then the solution of the
state differential equation is found to be
t

x ( t ) e At x (0) e A ( t ) B u () d
0

X(s) sI A x (0) sI A B U(s)

1

where we note that [sI-A]-1=(s), which is the Laplace transform

of (t)=eAt. The matrix exponential function (t) describes the
unforced response of the system and is called the fundamental or
state transition matrix.
t

x ( t ) ( t ) x (0) ( t ) B u () d
0

CONTROLLABILITY:
a system must be completely controllable and completely
observable to allow the flexibility to place all the closed-loop
system poles arbitrarily. The concepts of controllability and
observability were introduced by Kalman in the 1960s.

A system is completely controllable if there exists an unconstrained

control u(t) that can transfer any initial state x(t 0) to any other desired
location x(t) in a finite time, t0tT.

For the system

x Ax Bu
we can determine whether the system is controllable by examining
the algebraic condition

rank B AB A 2 B A n 1B n
The matrix A is an nxn matrix an B is an nx1 matrix. For multi input
systems, B can be nxm, where m is the number of inputs.
For a single-input, single-output system, the controllability matrix P c
is described in terms of A and B as

Pc B AB A 2 B A n 1B

which is nxn matrix. Therefore, if the determinant of Pc is nonzero,

the system is controllable.

Example:
Consider the system

0
0
a0

0
0
a0

1
0
a1

1 x

a 2

1 , B

a 2

1
0
a1

0
0 u
1

y 1 0 0 x 0 u

0
1
0

0 , AB 1 , A 2 B a 2
a 2
a 22 a 1
1

0
Pc B AB A 2 B 0
1

0
1
a2

1
a 2
a 22 a1

The determinant of Pc =1 and 0 , hence this system is controllable.

Example.
Consider a system represented by the two state equations

x 1 2 x1 u , x 2 3 x 2 d x1
The output of the system is y=x2. Determine the condition of
controllability.

2 0
1
x
x u , y 0 1 x 0 u

d 3
0
1
2 0 1
2
B and AB

0
d

3
0
d

1 2
Pc

The determinant of pc is equal to d, which

0
d

is nonzero only when d is nonzero.

Dorf and Bishop, Modern Control Systems

The controllability matrix Pc can be constructed in Matlab by using

ctrb command.
From two-mass system,

0
0

, A

500

500

20.5

50

20000 20000

0
1
0

8.2
Pc =

clc
clear
A=[0 0 1 0;0 0 0 1;-500 500 -20.5
0;20000 -20000 0 -8.2];
B=[0;0;50;0];
Pc=ctrb(A,B)
rank_Pc=rank(Pc)
det_Pc=det(Pc)

1.0e+007 *
0 0.0000 -0.0001 -0.0004
0
0
0 0.1000
0.0000 -0.0001 -0.0004 0.0594
0
0 0.1000 -2.8700
rank_Pc =
4
det_Pc =
-2.5000e+015

The system is
controllable.

OBSERVABILITY:
All the poles of the closed-loop system can be placed arbitrarily in
the complex plane if and only if the system is observable and
controllable. Observability refers to the ability to estimate a state
variable.

A system is completely observable if and only if there exists a finite time T

such that the initial state x(0) can be determined from the
observation history y(t) given the control u(t).

Consider the single-input, single-output system

x Ax Bu

and

y Cx

where C is a 1xn row vector, and x is an nx1 column vector. This

system is completely observable when the determinant of the
observability matrix P0 is nonzero.

The observability matrix, which is an nxn matrix, is written as

PO

C
C A

n 1
CA

Example:
Consider the previously given system

0
0
a0

1
0
a1

0
1 , C 1 0 0
a 2
Dorf and Bishop, Modern Control Systems

CA 0 1 0

, CA 2 0 0 1

Thus, we obtain

1 0 0
PO 0 1 0
0 0 1
The det P0=1, and the system is completely observable. Note that
determination of observability does not utility the B and C matrices.

Example: Consider the system given by

2 0
1
x
x
u

1 1
1

and

y 1 1 x

We can check the system controllability and observability using the P c

and P0 matrices.
From the system definition, we obtain

1
B
1

and

2
AB

Therefore, the controllability matrix is determined to be

2
1
Pc B AB

det Pc=0 and rank(Pc)=1. Thus, the system is not controllable.

Dorf and Bishop, Modern Control Systems

From the system definition, we obtain

C 1 1 and CA 1 1
Therefore, the observability matrix is determined to be

C 1 1
Po

CA
1
1

det PO=0 and rank(PO)=1. Thus, the system is not observable.

If we look again at the state model, we note that

y x1 x 2
However,

x 1 x 2 2x1 x 2 x1 u u x1 x 2

Thus, the system state variables do not depend on u, and the system
is not controllable. Similarly, the output (x 1+x2) depends on x1(0) plus
x2(0) and does not allow us to determine x1(0) and x2(0)
independently. Consequently, the system is not observable.
The observability matrix PO can be constructed in Matlab by using
obsv command.
From two-mass system,
Po =

clc
clear
A=[2 0;-1 1];
C=[1 1];
Po=obsv(A,C)
rank_Po=rank(Po)
det_Po=det(Po)

1
1

1
1

rank_Po =
1
det_Po =

The system is not

observable.

0
Dorf and Bishop, Modern Control Systems