IC8651 Advanced Control System Department of ICE 2022-23

UNIT I STATE VARIABLE ANALYSIS

Transfer function approach of system modelling provides final relation between output
variable and input variable. However, a system may have other internal variables of importance.
State variable representation takes into account of all such internal variables. Moreover, controller
design using classical methods, e.g., root locus or frequency domain method are limited to only LTI
systems, particularly SISO (single input single output) systems since for MIMO (multi input multi
output) systems controller design using classical approach becomes more complex. These
limitations of classical approach led to the development of state variable approach of system
modelling and control which formed a basis of modern control theory. State variable models are
basically time domain models where we are interested in the dynamics of some characterizing
variables called state variables which along with the input represent the state of a system at a given
time.
Drawbacks of transfer function analysis
1. It is defined under zero initial conditions
2. It is applicable to linear time invariant systems
3. It is restricted to Single Input Single Output (SISO) systems.
4. It does not provide any information regarding internal state of the system.
Advantages of state variable analysis
1. It can be applied to nonlinear system also.
2. It can be applied to time invariant systems also.
3. It can be applied to Multiple Input Multiple Output (MIMO) systems also.
4. Its gives idea about the internal state of the system.

State space formulation

State: The state of a dynamic system is a minimal set of variables called state variables such that the
knowledge of these variables at time 𝑡 = 𝑡0 together with the knowledge of inputs for ≥ 𝑡0 ,
completely determines the behaviour of the system for 𝑡 > 𝑡0 . A set of variables which describes the
system at any time instant are called state variables.
State vector: If 𝑛 state variables are needed to completely describe the behaviour of a given system,
then these 𝑛 state variables can be considered the components of a vector 𝑋. Such a vector is called
as state vector.
State space: The dimensional space whose co-ordinate axes consists of the 𝑥1 axis, 𝑥2 axis,……..
𝑥𝑛 axis, where 𝑥1 , 𝑥2 , … . . 𝑥𝑛 are state variables is called a state space

In general a system consists of inputs, outputs and state variables.

State variables = 𝑥1 (𝑡), 𝑥2 (𝑡), … … … . . 𝑥𝑛 (𝑡)
Input variables = 𝑢1 (𝑡), 𝑢2 (𝑡), … … … . . 𝑢𝑚 (𝑡)
Output variables = 𝑦1 (𝑡), 𝑦2 (𝑡), … … … . . 𝑦𝑝 (𝑡)

1 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

The different variables may be represented by the vectors as shown below

𝑢1 (𝑡) 𝑦1 (𝑡) 𝑥1 (𝑡)
𝑢2 (𝑡) 𝑦2 (𝑡) 𝑥2 (𝑡)
Input vector 𝑈(𝑡) = ; ; Output vector 𝑌(𝑡) = ; ; State variable vector 𝑋(𝑡) = ;
; ; ;
[𝑢𝑚 (𝑡)] [𝑦𝑝 (𝑡)] [𝑥𝑛 (𝑡)]
State Equations: The state variable representation can be arranged in the form of number of first
order differential equations.
𝑑𝑥1
= 𝑥̇ 1 = 𝑓1 (𝑥1 , 𝑥2 , … … … . . 𝑥𝑛 ; 𝑢1 , 𝑢2 , … … … . . 𝑢𝑚 )
𝑑𝑡
𝑑𝑥2
= 𝑥̇ 2 = 𝑓2 (𝑥1 , 𝑥2 , … … … . . 𝑥𝑛 ; 𝑢1 , 𝑢2 , … … … . . 𝑢𝑚 )
𝑑𝑡
;
𝑑𝑥𝑛
= 𝑥̇ 𝑛 = 𝑓𝑛 (𝑥1 , 𝑥2 , … … … . . 𝑥𝑛 ; 𝑢1 , 𝑢2 , … … … . . 𝑢𝑚 )
𝑑𝑡
These number of differential equations may be written in vector notations as
𝑋̇(𝑡) = 𝑓(𝑋(𝑡), 𝑈(𝑡)
The set of all possible values which the input vector 𝑈(𝑡) can have at time 𝑡 forms the input space of
the system. Similarly the set of all possible values which the output vector 𝑌(𝑡) can assume at time
𝑡 forms output space of the system and the set of all possible values which the output vector 𝑋(𝑡)
can assume at time 𝑡 forms state space of the system.
State model of linear system: The state model of a system consists of the state equation and output
equation. The state equation of a system is a function of the state variables and inputs. For an LTI
system the first derivative of state variables can be expressed as a linear combination of sate variables
and inputs.
𝑥̇ 1 = 𝑎11 𝑥1 + 𝑎12 𝑥2 + , … … … . . 𝑎1𝑛 𝑥𝑛 + 𝑏11 𝑢1 + 𝑏12 𝑢2 +, … … … . . 𝑏1𝑚 𝑢𝑚
𝑥̇ 2 = 𝑎21 𝑥1 + 𝑎22 𝑥2 + , … … … . . 𝑎2𝑛 𝑥𝑛 + 𝑏21 𝑢1 + 𝑏22 𝑢2 +, … … … . . 𝑏2𝑚 𝑢𝑚
;
𝑥̇ 𝑛 = 𝑎𝑛1 𝑥1 + 𝑎𝑛2 𝑥2 + , … … … . . 𝑎𝑛𝑛 𝑥𝑛 + 𝑏𝑛1 𝑢1 + 𝑏𝑛2 𝑢2 +, … … … . . 𝑏𝑛𝑚 𝑢𝑚
Where the coefficients 𝑎𝑖𝑗 and 𝑏𝑖𝑗 are constants. In the matrix form this can be expressed as,
𝑥̇ 1 𝑎11 𝑎12 … … … … 𝑎1𝑛 𝑥1 𝑏11 𝑏12 … … … … 𝑏1𝑚 𝑢1
𝑥̇ 2 𝑎21 𝑎22 … … … … 𝑎2𝑛 𝑥2 𝑏21 𝑎22 … … … … 𝑎2𝑚 𝑢2
𝑥̇ 3 = 𝑎31 𝑎32 … … … … 𝑎3𝑛 𝑥3 + 𝑏31 𝑏32 … … … … 𝑏3𝑚 𝑢3
; ; ; …………….. ; ; ; …………….. ;
[𝑥̇ 𝑛 ] [𝑎𝑛1 𝑎𝑛2 … … … … 𝑎𝑛𝑛 ] [𝑥𝑛 ] [𝑏𝑛1 𝑏𝑛2 … … … … 𝑏𝑛𝑚 ] [𝑢𝑚 ]
This matrix equation can also be written as, 𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡)
Where, 𝑋(𝑡) = State vector of order (𝑛 × 1)
𝑈(𝑡) = Input vector of order (𝑚 × 1)
𝐴 = System matrix of order (𝑛 × 𝑛)
𝐵 = Input matrix of order (𝑛 × 𝑚)
This equation is called the state equation of LTI system. The output at any time are function of state
variables and inputs.
Output vector, 𝑌(𝑡) = 𝑓(𝑋(𝑡), 𝑈(𝑡))
Hence the output variables can be expressed as a linear combination of state variables and inputs,
𝑦1 = 𝑐11 𝑥1 + 𝑐12 𝑥2 + , … … … . . 𝑐1𝑛 𝑥𝑛 + 𝑑11 𝑢1 + 𝑑12 𝑢2 +, … … … . . 𝑑1𝑚 𝑢𝑚
𝑦2 = 𝑐21 𝑥1 + 𝑐22 𝑥2 + , … … … . . 𝑐2𝑛 𝑥𝑛 + 𝑑21 𝑢1 + 𝑑22 𝑢2 +, … … … . . 𝑑2𝑚 𝑢𝑚
;
𝑦𝑛 = 𝑐𝑝1 𝑥1 + 𝑐𝑝2 𝑥2 + , … … … . . 𝑐𝑝𝑛 𝑥𝑛 + 𝑑𝑝1 𝑢1 + 𝑑𝑝2 𝑢2 +, … … … . . 𝑑𝑝𝑚 𝑢𝑚

2 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Where the coefficients 𝑐𝑖𝑗 and 𝑑𝑖𝑗 are constants. In the matrix form this can be expressed as,
𝑦1 (𝑡) 𝑐11 𝑐12 … … … … 𝑐1𝑛 𝑥1 𝑑11 𝑑12 … … … … 𝑑1𝑚 𝑢1
𝑦2 (𝑡) 𝑐21 𝑐22 … … … … 𝑐2𝑛 𝑥2 𝑑21 𝑑22 … … … … 𝑑2𝑚 𝑢2
𝑦3 (𝑡) = 𝑐31 𝑐32 … … … … 𝑐3𝑛 𝑥3 + 𝑑31 𝑑32 … … … … 𝑑3𝑚 𝑢3
; ; ; …………….. ; ; ; …………….. ;
(𝑡)
[𝑦𝑝 ] [𝑐 𝑝1 𝑐𝑝2 … … … … 𝑐𝑝𝑛 ] [𝑥𝑛 ] 𝑑
[ 𝑝1 𝑑 𝑝2 … … … … 𝑑 𝑢
𝑝𝑚 ] 𝑛𝑚 ]
[
This matrix equation can also be written as, 𝑌(𝑡) = 𝐶𝑋(𝑡) + 𝐷𝑈(𝑡)
Where, 𝑋(𝑡) = State vector of order (𝑛 × 1)
U(𝑡) = Input vector of order (𝑚 × 1)
𝑌(𝑡) = Output vector of order (𝑝 × 1)
𝐶 = Output matrix of order (𝑝 × 𝑛)
𝐷 = Transmission matrix of order (𝑝 × 𝑚)
This equation is called the output equation of LTI system. The state model of a system consists of
state equation and output equation.
𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡)………… State equation
𝑌(𝑡) = 𝐶𝑋(𝑡) + 𝐷𝑈(𝑡) ………... Output equation

State diagram: The pictorial representation of the state model of the system is called state diagram.
It can be either Block diagram or signal flow graph form. The state diagram describes the
relationships among the state variables and provides physical interpretations of the state variables
time domain state diagram may be obtained directly from the differential equation governing the
system and this diagram can be used for simulation of the system in computer. The s-domain state
diagram can be obtained from the transfer function of the system. The state diagram of a state model
is constructed using three basic elements Scalar, Adder and integrator.
Scalar: the scalar is used to multiply a signal by a constant. The input signal is multiplied by the
scalar to give the output.
Adder: The adder is used to add two or more signals. The output of the adder is the sum of incoming
signals.
Integrator: the integrator is used to integrate the signal. They are used to integrate the derivatives
of the state variables to get the state variable the initial condition of the state variable can be added
by using an adder after integrator.
Elements of Block diagram and Signal flow graph
Block diagram Signal flow graph
Element
Time domain s-domain Time domain s-domain
Scalar

Adder

Integrator

The state model of LTI system is given by,

𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡)………… State equation
𝑌(𝑡) = 𝐶𝑋(𝑡) + 𝐷𝑈(𝑡) ………... Output equation

3 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

The time domain block diagram representation and time domain signal flow graph representation
of the state model is given by,

STATE SPACE REPRESENTATION USING PHYSICAL VARIABLES

In state space modelling the choice of state variable is arbitrary. One of the possible choice is physical
variables. The physical variables of electrical systems are current or voltage and in mechanical
systems displacement, velocity and acceleration. The advantages of choosing physical variables of
the systems as state variables are;
1. State variables can be utilized for the purpose of feedback.
2. The implementation of design with state variable feedback becomes straight forward.
3. The solution of state equation give time variation of variables which have direct relevance to
the physical system.
The drawback of choosing physical quantities as state variables is that the solution of state equation
may become a difficult task.
Electrical System
The basic model of an electrical system can be obtained by using the fundamental elements Resistor,
Capacitor and Inductor. Using these elements the electrical network or equivalent circuit of the
system is drawn. Then the differential equations governing the electrical systems can be formed by
writing Kirchoff's current law equations by choosing various nodes in the network or Kirchoff's
voltage law by choosing various closed path in the network. The current-voltage relation of the basic
elements R, L and C are given in table.
Voltage across the Current through the
Element
element element
𝑣(𝑡)
𝑣(𝑡) = 𝑅 𝑖(𝑡) 𝑖(𝑡) =
𝑅
𝑑 1
𝑣(𝑡) = 𝐿 𝑖(𝑡) 𝑖(𝑡) = ∫ 𝑣(𝑡) 𝑑𝑡
𝑑𝑡 𝐿
1 𝑑
𝑣(𝑡) = ∫ 𝑖(𝑡) 𝑑𝑡 𝑖(𝑡) = 𝐶 𝑣(𝑡)
𝐶 𝑑𝑡

A minimal number of state variables are chosen for obtaining the state model of the system. The best
choice of state variables in electrical system are currents and voltages in energy storage elements.
The energy storage elements are inductance and capacitance. The physical variables in the
differential equations are replaced by state variables and the equations are rearranged as first order
differential equations. These set of first order equations constitutes the state equation of the system.
The inputs to the system are exciting voltage sources or current sources. The outputs in electrical
4 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

system are usually voltages or currents in energy dissipating element. The resistance is energy
dissipating element in electrical network. In general the output variables can be any voltage or
current in the network.

Mechanical Translational System

The basic model of mechanical translational system can be obtained by using three basic elements
mass, spring and dash-pot. When a force is applied to a mechanical translational system, it is
opposed by opposing forces due to mass, friction and elasticity of the system. The forces acting on a
body are governed by Newton's second law of motion. The differential equations governing the
system are obtained by writing force balance equations at various nodes in the system. A node is a
meeting point of elements. The table shows the force balance equations of idealized elements.
List of symbols used in mechanical translational system
𝑦 = Displacement, 𝑚
𝑑𝑦
𝑣= 𝑑𝑡
= Velocity, 𝑚/𝑠𝑒𝑐
𝑑𝑣 𝑑2 𝑦
𝑎= = = Acceleration, 𝑚/𝑠𝑒𝑐 2
𝑑𝑡 𝑑𝑡 2
𝑓 = Applied force, 𝑁(𝑁𝑒𝑤𝑡𝑜𝑛)
𝑓𝑚 = Opposing force offered by mass of the body, 𝑁
𝑓𝑏 = Opposing force offered by the friction of the body (Dash-pot), 𝑁
𝑓𝑘 = Opposing force offered by the elasticity of the body (Spring), 𝑁
𝑀 = Mass, 𝐾𝑔
𝐵 = Viscous friction coefficient, 𝑁/(𝑚/𝑠𝑒𝑐)
𝐾 = Stiffness of spring, 𝑁/𝑚
Force balance equations of idealized elements
Mechanical Translational system
Element Force balance equation

𝑑2 𝑦
𝑓 = 𝑓𝑚 = 𝑀
𝑑𝑡 2

𝑑𝑦
𝑓 = 𝑓𝑏 = 𝐵
𝑑𝑡

𝑑
𝑓 = 𝑓𝑏 = 𝐵 (𝑦 − 𝑦2 )
𝑑𝑡 1

𝑓 = 𝑓𝑘 = 𝐾 𝑦

𝑓 = 𝑓𝑘 = 𝐾(𝑦1 − 𝑦2 )

Mechanical rotational system

The basic model of mechanical translational system can be obtained by using three basic elements
moment of inertia of mass, rotational dash-pot and rotational spring. When a torque is applied to a
mechanical rotational system, it is opposed by opposing torques due to moment of inertia, friction
and elasticity of the system. The torques acting on a body are governed by Newton's second law of
motion. The differential equations governing the system are obtained by writing torque balance
5 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

equations at various nodes in the system. A node is a meeting point of elements. The table shows
the torque balance equations of idealized elements.
List of symbols used in mechanical rotational system
𝜃 = Displacement, 𝑟𝑎𝑑
𝑑𝜃
𝑣= = Angular velocity, 𝑟𝑎𝑑/𝑠𝑒𝑐
𝑑𝑡
𝑑𝑣 𝑑2 𝜃
𝑎= 𝑑𝑡
= 𝑑𝑡 2
= Angular acceleration, 𝑟𝑎𝑑/𝑠𝑒𝑐 2
𝑇 = Applied torque, 𝑁 − 𝑚
𝑇𝑗 = Opposing torque offered by moment of inertia of mass, 𝑁 − 𝑚
𝑇𝑏 = Opposing torque offered by the rotational Dash-pot, 𝑁 − 𝑚
𝑇𝑘 = Opposing torque offered by the rotational Spring, 𝑁 − 𝑚
𝐽 = Moment of Inertia, 𝐾𝑔 − 𝑚2 /𝑟𝑎𝑑
𝐵 = Rotational frictional coefficient, 𝑁 − 𝑚/(𝑟𝑎𝑑/𝑠𝑒𝑐)
𝐾 = Stiffness of spring, 𝑁 − 𝑚/𝑟𝑎𝑑

Torque balance equations of idealized elements

Mechanical Rotational System
Element Torque balance equation
𝑑2 𝜃
𝑇 = 𝑇𝑗 = 𝐽
𝑑𝑡 2
𝑑𝜃
𝑇 = 𝑇𝑏 = 𝐵
𝑑𝑡
𝑑
𝑇 = 𝑇𝑏 = 𝐵 (𝜃 − 𝜃2 )
𝑑𝑡 1

𝑇 = 𝑇𝑘 = 𝐾 𝜃

𝑇 = 𝑇𝑘 = 𝐾(𝜃1 − 𝜃2 )

Guidelines to form the state model of mechanical translational and rotational systems
Sl.No Mechanical translational systems Mechanical rotational systems
1. Identify the nodes. Nodes are the meeting Identify the nodes. Nodes are the meeting
point of elements. In general masses are point of elements. In general moment of
assumed as nodes. Inertia of masses are assumed as nodes.
2. Assign a displacement to each node and Assign an angular displacement to each node
draw a free body diagram for each node. and draw a free body diagram for each node.
The free body diagram is obtained by The free body diagram is obtained by drawing
drawing each mass of node separately and each mass of node separately and then
then marking all the forces acting on it. marking all the torques acting on it.
3. In the free body diagram, the opposing In the free body diagram, the opposing
forces due to mass, spring and dash-pot are torques due to moment of inertia, spring and
always act in a direction opposite to dash-pot are always act in a direction opposite
applied force. The displacement, velocity to applied torque. The angular displacement,
and acceleration will be in the direction of velocity and acceleration will be in the

6 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

applied force or in the direction opposite to direction of applied torque or in the direction
that of opposing force. opposite to that of opposing torque.
4. For each free body diagram write one For each free body diagram write one
differential equation by equating the sum differential equation by equating the sum of
of applied force to the sum of opposing applied torque to the sum of opposing
forces. torques.
5. Choose a minimum number of state Choose a minimum number of state variables.
variables. The choice of variables are The choice of variables are angular
displacement, velocity and acceleration. displacement, velocity and acceleration.
6. The physical variables in differential The physical variables in differential
equations are replaced by state variables equations are replaced by state variables and
and the equations are rearranged as first the equations are rearranged as first order
order differential equations. These set of differential equations. These set of first order
first order equations constitute the state equations constitute the state equation of the
equation of the system. system.
7. The inputs are the applied forces and the The inputs are the applied torques and the
outputs are the displacement, velocity or outputs are the angular displacement, velocity
acceleration of the desired nodes. or acceleration of the desired nodes.

Example 1: Construct the state model of mechanical system shown in

figure.

Solution:
The free body diagram of 𝑀1 is shown in figure.
𝑑2 𝑦1
𝑓𝑚1 = 𝑀1
𝑑𝑡 2
𝑑
𝑓𝑏1 = 𝐵1 (𝑦1 − 𝑦2 )
𝑑𝑡
𝑓𝑘1 = 𝐾1 (𝑦1 − 𝑦2 )

By Newton’s second law, the force balance equation at node 𝑀1 is

𝑓(𝑡) = 𝑓𝑚1 + 𝑓𝑏1 + 𝑓𝑘1
𝑑2 𝑦1 𝑑
𝑓(𝑡) = 𝑀1 2 + 𝐵1 (𝑦1 − 𝑦2 ) + 𝐾1 (𝑦1 − 𝑦2 )
𝑑𝑡 𝑑𝑡
𝑑2 𝑦1 𝑑𝑦1 𝑑𝑦2
𝑓(𝑡) = 𝑀1 2 + 𝐵1 − 𝐵1 + 𝐾1 𝑦1 − 𝐾1 𝑦2 (1)
𝑑𝑡 𝑑𝑡 𝑑𝑡
The free body diagram of 𝑀2 is shown in figure.

.
By Newton’s second law, the force balance equation at node 𝑀2 is
𝑓𝑚2 + 𝑓𝑏2 + 𝑓𝑏1 + 𝑓𝑘2 + 𝑓𝑘1 = 0

7 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

𝑑2 𝑦2 𝑑𝑦2 𝑑
𝑀2 2
+ 𝐵2 + 𝐵1 (𝑦2 − 𝑦1 ) + 𝐾2 𝑦2 + 𝐾1 (𝑦2 − 𝑦1 ) = 0
𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑑2 𝑦2 𝑑𝑦2 𝑑𝑦2 𝑑𝑦1
𝑀2 2
+ 𝐵2 + 𝐵1 − 𝐵1 + 𝐾2 𝑦2 + 𝐾1 𝑦2 − 𝐾1 𝑦1 = 0 (2)
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡
Let us choose four state variables 𝑥1 , 𝑥2 , 𝑥3 and 𝑥4 . Also let the input 𝑓(𝑡) = 𝑢. The state variables are
related to physical variables as follows.
𝑑𝑦1 𝑑𝑦2 𝑑2 𝑦1 𝑑2 𝑦2
𝑥1 = 𝑦1 ; 𝑥2 = 𝑦2 ; 𝑥3 = ; 𝑥4 = ; 𝑥̇ 3 = ; 𝑥̇ 4 = (3)
𝑑𝑡 𝑑𝑡 𝑑𝑡 2 𝑑𝑡 2
On substituting (3) and 𝑓(𝑡) = 𝑢 in equation (1), we get,
𝑢 = 𝑀1 𝑥̇ 3 + 𝐵1 𝑥3 − 𝐵1 𝑥4 + 𝐾1 𝑥1 − 𝐾1 𝑥2
𝑀1 𝑥̇ 3 = −𝐵1 𝑥3 + 𝐵1 𝑥4 − 𝐾1 𝑥1 + 𝐾1 𝑥2 + 𝑢
𝐾1 𝐾1 𝐵1 𝐵1 1
𝑥̇ 3 = − 𝑥1 + 𝑥2 − 𝑥3 + 𝑥4 + 𝑢 (4)
𝑀1 𝑀1 𝑀1 𝑀1 𝑀1
On substituting (3) in equation (2),
𝑀2 𝑥̇ 4 + 𝐵2 𝑥4 + 𝐵1 𝑥4 − 𝐵1 𝑥3 + 𝐾2 𝑥2 + 𝐾1 𝑥2 − 𝐾1 𝑥1 = 0
𝑀2 𝑥̇ 4 = −𝐵2 𝑥4 − 𝐵1 𝑥4 + 𝐵1 𝑥3 − 𝐾2 𝑥2 − 𝐾1 𝑥2 + 𝐾1 𝑥1
𝐾1 (𝐾1 + 𝐾2 ) 𝐵1 (𝐵1 + 𝐵2 )
𝑥̇ 4 = 𝑥1 − 𝑥2 + 𝑥3 − 𝑥4 (5)
𝑀2 𝑀2 𝑀2 𝑀2
From equation (3), the state variable
𝑥1 = 𝑦1
𝑑𝑥1 𝑑𝑦1
Differentiating this w.r.t 𝑡, =
𝑑𝑡 𝑑𝑡
𝑑𝑥1 𝑑𝑦1
Let = 𝑥̇ 1 and = 𝑥3 ; ∴ 𝑥̇ 1 = 𝑥3 (6)
𝑑𝑡 𝑑𝑡
From equation (3), the state variable
𝑥2 = 𝑦2
𝑑𝑥2 𝑑𝑦2
Differentiating this w.r.t 𝑡, 𝑑𝑡
= 𝑑𝑡
𝑑𝑥2 𝑑𝑦2
Let 𝑑𝑡
= 𝑥̇ 2 and 𝑑𝑡
= 𝑥4 ; ∴ 𝑥̇ 2 = 𝑥4 (7)
The equations (4), (5), (6) and (7) are state equations of the mechanical system.
𝑥̇ 1 = 𝑥3
𝑥̇ 2 = 𝑥4
𝐾1 𝐾1 𝐵1 𝐵1 1
𝑥̇ 3 = − 𝑥1 + 𝑥2 − 𝑥3 + 𝑥4 + 𝑢
𝑀1 𝑀1 𝑀1 𝑀1 𝑀1
𝐾1 (𝐾1 + 𝐾2 ) 𝐵1 (𝐵1 + 𝐵2 )
𝑥̇ 4 = 𝑥1 − 𝑥2 + 𝑥3 − 𝑥4
𝑀2 𝑀2 𝑀2 𝑀2
On arranging the state equation in the matrix form,
0 0 1 0
𝑥̇ 1 𝑥1 0
0 0 0 1
𝑥̇ 𝐾 𝐾1 𝐵1 𝐵1 𝑥2 0
[ 2 ] = − 𝑀1 − [ 𝑥3 ] + 1 [𝑢] (8)
𝑥̇ 3 1 𝑀2 𝑀1 𝑀1
𝑀1
− 1𝑀 2 ] 𝑥4
𝑥̇ 4 𝐾1 (𝐾1 +𝐾2 ) 𝐵1 (𝐵 +𝐵 )
[0]
[ 𝑀2 − 𝑀2 𝑀2 2
Let the displacement 𝑦1 and 𝑦2 be the outputs of the system
𝑦1 = 𝑥1 and 𝑦2 = 𝑥2
The output equation in matrix form is given by,
𝑥1
𝑦1 1 0 0 0 2 𝑥
[𝑦 ] = [ ][ ] (9)
2 0 1 0 0 𝑥3
𝑥4
The state equation (8) and the output equation (9) together called the state model of the system.

8 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Example 2: Obtain the state model of the mechanical system shown by choosing a minimum of
three state variables.

Solution
Let the state variables be 𝑥1 , 𝑥2 and , 𝑥3 and they are related to physical variables as shown below
𝑑𝑦1
𝑥1 = 𝑦1 ; 𝑥2 = 𝑦2 ; 𝑥3 =
𝑑𝑡
Free body diagram of mass 𝑀 is shown in figure
By Newton’s second law, the force equation at node 𝑀 is
𝑓𝑚 + 𝑓𝑘1 + 𝑓𝑘2 = 0
𝑑2 𝑦1
𝑀 2 + 𝐾1 𝑦1 + 𝐾2 (𝑦2 − 𝑦1 ) = 0
𝑑𝑡
𝑑 2 𝑦1
𝑀 𝑑𝑡 2
+ 𝐾1 𝑦1 + 𝐾2 𝑦1 − 𝐾2 𝑦2 = 0 (1)
𝑑 2 𝑦1
Put 𝑦1 = 𝑥1 ; 𝑦2 = 𝑥2 ; 𝑥̇ 3 = in equation (1)
𝑑𝑡 2
𝑀𝑥̇ 3 + 𝐾1 𝑥1 + 𝐾2 𝑥1 − 𝐾2 𝑥2 = 0
𝑀1 𝑥̇ 3 + (𝐾1 + 𝐾2 )𝑥1 − 𝐾2 𝑥2 = 0
𝐾1 +𝐾2 𝐾2
𝑥̇ 3 = − 𝑀
𝑥1 + 𝑀 2
𝑥 (2)
The free body diagram of node 2 (meeting point of 𝐾2 and 𝐵) is shown in figure
By Newton’s second law, the force equation at node 2 is
𝑓𝑏 + 𝑓𝑘2 = 0
𝑑𝑦2
𝐵 + 𝐾2 (𝑦2 − 𝑦1 ) = 0
𝑑𝑡
𝑑𝑦
𝐵 𝑑𝑡2 + 𝐾2 𝑦2 − 𝐾2 𝑦1 = 0 (3)
𝑑𝑦2
Put 𝑦1 = 𝑥1 ; 𝑦2 = 𝑥2 ; 𝑥̇ 2 = 𝑑𝑡
in equation (3)
𝐵𝑥̇ 2 + 𝐾2 𝑥2 − 𝐾2 𝑥1 = 0
𝐵𝑥̇ 2 = 𝐾2 𝑥1 − 𝐾2 𝑥2
𝐾2 𝐾2
𝑥̇ 2 = 𝐵 1
𝑥 − 𝐵 2
𝑥 (4)
′
The state variable 𝑥1 = 𝑦1 on differentiating this expression with respect to ′𝑡
𝑑𝑥1 𝑑𝑦1
= = 𝑥3 ∴ 𝑥̇ 1 = 𝑥3 (5)
𝑑𝑡 𝑑𝑡
The equations (2), (4) and (5) are state equations of the mechanical system.
𝑥̇ 1 = 𝑥3
𝐾2 𝐾2
𝑥̇ 2 = 𝑥1 − 𝑥2
𝐵 𝐵
𝐾1 + 𝐾2 𝐾2
𝑥̇ 3 = − 𝑥1 + 𝑥2
𝑀 𝑀
On arranging the state equation in the matrix form,
0 0 1 𝑥
𝑥̇ 1 𝐾2 𝐾2 1
[𝑥̇ 2 ] = [ 𝐵 − 𝐵 0] [𝑥2 ] (6)
0 𝑥3
𝑥̇ 3 𝐾 +𝐾 𝐾2
− 1 2 𝑀 𝑀
If the desired outputs are 𝑦1 and 𝑦2 then 𝑦1 = 𝑥1 and 𝑦2 = 𝑥2
9 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

The output equation is the matrix form is given by,

𝑦1 1 0 𝑥1
[𝑦 ] = [ ][ ] (7)
2 0 1 𝑥2
The state equation (5) and output equation (6) together constitute the state model of the system.

STATE SPACE REPRESENTATION USING PHASE VARIABLES

The phase variables are defined as those particular state variables which are obtained from one of
the system variables and its derivatives. Usually the variable used is the system output and the
remaining state variables are then derivatives of the output. The state model using phase variables
can be easily determined if the system model is already known in the differential equation or transfer
function form.
Advantages: the state space model can be directly formed by inspection from the differential
equations governing the system. The phase variables provides a link between the transfer function
design approach and time domain design approach.
Disadvantages: the phase variables are not physical variables of the system and therefore are not
available for measurement and control purposes.

Method I Procedure:
Consider the following 𝑛𝑡ℎ order linear differential equation relating the output 𝑦(𝑡) to the input
𝑢(𝑡) of a system

By choosing the output 𝑦 and their derivatives as state variables

𝑥1 = 𝑦
𝑥2 = 𝑦̇
𝑥3 = 𝑦̈
;
𝑥𝑛 = 𝑦̇
On substituting the state variables in the differential equation governing the system,
𝑥̇ 𝑛 + 𝑎1 𝑥𝑛 + 𝑎2 𝑥𝑛−1 + 𝑎3 𝑥𝑛−2 +, … … … … … + 𝑎𝑛−2 𝑥3 + 𝑎𝑛−1 𝑥2 + 𝑎𝑛 𝑥1 = 𝑏 𝑢
𝑥̇ 𝑛 = −𝑎𝑛 𝑥1 − 𝑎𝑛−1 𝑥2 − 𝑎𝑛−2 𝑥3 −, … … … … … − 𝑎2 𝑥𝑛−1 + 𝑎1 𝑥𝑛 + 𝑏 𝑢
The state equations of the system are
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 3 = 𝑥4
;
𝑥̇ 𝑛−1 = 𝑥𝑛
𝑥̇ 𝑛 = −𝑎𝑛 𝑥1 − 𝑎𝑛−1 𝑥2 − 𝑎𝑛−2 𝑥3 −, … … … … … − 𝑎2 𝑥𝑛−1 + 𝑎1 𝑥𝑛 + 𝑏 𝑢
On arranging the above equations in the matrix form,
𝑥̇ 1 0 1 0 0 ……… 0 𝑥1 0
𝑥̇ 2 0 0 1 0 ……… 0 𝑥2 0
𝑥̇ 3 0 0 0 1 ……… 0 𝑥3 0
;
=
; ; ; ; ; ; ; + ; [𝑢]
𝑥̇ 𝑛−1 0 0 0 0 ……… 1 𝑥𝑛−1 0
[ 𝑥̇ 𝑛 ] [−𝑎 𝑛 −𝑎 𝑛−1 −𝑎 𝑛−2 −𝑎 𝑛−3 … … … −𝑎 1 ] [ 𝑥𝑛 ] [𝑏]
Or 𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡)
Here matrix 𝐴 (system matrix) has a very special form. It has all 1’s in the upper off-diagonal, its last
row is comprised of the negative of the coefficients of the original differential equation and all other
elements are zero. This form of matrix 𝐴 is known as Bush form (or) companion form. Also 𝐵 matrix
10 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

has the specialty that all the elements except the last element are zero. The output being 𝑦 = 𝑥1, the
output equation is given by,
𝑥1
𝑥2
𝑥3
𝑦 = [1 0 0 … … … 0] ; or 𝑌 = 𝐶 𝑋
𝑥𝑛−1
[ 𝑥𝑛 ]
The advantage in using phase variables for state space modelling is that the system state model can
be written directly by inspection from the differential equation governing the system.

Method II - Procedure
Consider the following 𝑛𝑡ℎ order linear differential equation relating the output 𝑦(𝑡) to the input
𝑢(𝑡) of a system
Let 𝑛 = 𝑚 = 3
∴ 𝑦⃛ + 𝑎1 𝑦̈ + 𝑎2 𝑦̇ + 𝑎3 𝑦 = 𝑏0 𝑢
⃛ + 𝑏1 𝑢̈ + 𝑏2 𝑢̇ + 𝑏3 𝑢 (1)

On taking laplace transform of (1) with zero initial conditions,

𝑆 3 𝑌(𝑆) + 𝑎1 𝑆 2 𝑌(𝑆) + 𝑎2 𝑆𝑌(𝑆) + 𝑎3 𝑌(𝑆) = 𝑏0 𝑆 3 𝑈(𝑆) + 𝑏1 𝑆 2 𝑈(𝑆) + 𝑏2 𝑆𝑈(𝑆) + 𝑏3 𝑈(𝑆) (2)

𝑌(𝑆) 𝑏0 𝑆 3 +𝑏1 𝑆 2 +𝑏2 𝑆+𝑏3

𝑈(𝑆)
= 𝑆 3 +𝑎1 𝑆 2 +𝑎2 𝑆+𝑎3

𝑏 𝑏 𝑏 𝑏 𝑏 𝑏
𝑆 3 (𝑏0 + 1+ 22+ 33 ) 𝑏0 + 1+ 22+ 33
= 𝑎
𝑆 𝑆
𝑎
𝑆
𝑎 = 𝑎
𝑆
𝑎
𝑆
𝑎
𝑆
(3)
𝑆 3 (1+ 1 + 22 + 33 ) 1−(− 1 − 22 − 33 )
𝑆 𝑆 𝑆 𝑆 𝑆 𝑆
From the mason’s gain formula, the transfer function of the system is given by

1
𝑇(𝑆) = ∑𝐾 𝑃𝐾 ∆𝐾
∆
Where, 𝑃𝐾 =path gain of 𝐾 𝑡ℎ forward path
∆=1-(sum of loop gain of all individual loops)
+(sum of gain products of all possible combinations of two non-touching loops)
-(sum of gain products of all possible combinations of three non-touching loops)
+….
∆𝐾 = ∆ for that part of the graph which is not touching 𝐾 𝑡ℎ forward path
The transfer function of a system with four forward paths and with three feedback loops is given
by,
𝑃 +𝑃2 +𝑃3 +𝑃4
1
𝑇(𝑆) = 1−(𝑃 (4)
11 +𝑃12 +𝑃13 )
On comparing (3) and (4),
𝑏1 𝑏 𝑏
𝑃1 = 𝑏2 ; 𝑃2 = 𝑆
; 𝑃3 = 𝑆22 and 𝑃4 = 𝑆33
𝑎1 𝑎2 𝑎3
𝑃11 = − ; 𝑃12 = − and 𝑃13 = −
𝑆 𝑆2 𝑆3
Hence the system represented by the transfer function as that of the equation (3), a signal flow graph
can be constructed. The signal flow graph is constructed such that all ∆𝐾 = 1 and all loops are
touching. Let us assign state variables at the output of each integrator in the signal flow graph. Hence
at the input of each integrator, the first derivative of the state variable will be available. The state
equations are formed by summing all the incoming signals to the nodes whose values corresponds
to first derivative of state variables.

11 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

By summing up the incoming signals at node 𝑥̇ 1 ,

𝑥̇ 1 = −𝑎1 (𝑥1 + 𝑏0 𝑢) + 𝑥2 + 𝑏1 𝑢
𝑥̇ 1 = −𝑎1 𝑥1 + 𝑥2 + (𝑏1 − 𝑎1 𝑏0 )𝑢 (5)

By summing up the incoming signals at node 𝑥̇ 2 ,

𝑥̇ 2 = −𝑎2 (𝑥1 + 𝑏0 𝑢) + 𝑥3 + 𝑏2 𝑢
𝑥̇ 2 = −𝑎2 𝑥1 + 𝑥3 + (𝑏2 − 𝑎2 𝑏0 )𝑢 (6)

By summing up the incoming signals at node 𝑥̇ 3 ,

𝑥̇ 3 = −𝑎3 (𝑥1 + 𝑏0 𝑢) + 𝑏3 𝑢
𝑥̇ 3 = −𝑎3 𝑥1 + (𝑏3 − 𝑎3 𝑏0 )𝑢 (7)

The output equation is given by the sum of incoming signals to the output node.
∴ 𝑦 = 𝑥1 + 𝑏0 𝑢 (8)
On arranging the state equations and the output equations in the matrix form,

𝑥̇ 1 −𝑎1 1 0 𝑥1 𝑏1 − 𝑎1 𝑏0
[𝑥̇ 2 ] = [−𝑎2 𝑥
0 1] [ 2 ] + [𝑏2 − 𝑎2 𝑏0 ] [𝑢] (9)
𝑥̇ 3 −𝑎3 0 0 𝑥3 𝑏3 − 𝑎3 𝑏0
𝑥1
[ ] 𝑥
𝑦 = 1 0 0 [ 2 ] + 𝑏0 𝑢 (10)
𝑥3
𝑡ℎ𝑒
𝑡ℎ
The above results can be generalized for an 𝑛 order differential equation and the general state
model is given by,
𝑥̇ 1 −𝑎1 1 0 0 … … … 0 𝑥1 𝑏1 − 𝑎1 𝑏0
𝑥̇ 2 −𝑎2 0 1 0 … … … 0 𝑥2 𝑏2 − 𝑎2 𝑏0
𝑥̇ 3 −𝑎3 0 0 1 … … … 0 𝑥3 𝑏3 − 𝑎3 𝑏0
= ; + [𝑢]
; ; ; ; ; ; ; ;
𝑥̇ 𝑛−1 −𝑎𝑛−1 0 0 0 … … … 1 𝑥𝑛−1 𝑏𝑛−1 − 𝑎𝑛−1 𝑏0
[ 𝑥̇ 𝑛 ] [ −𝑎𝑛 0 0 0 … … … 0] [ 𝑥𝑛 ] [ 𝑏𝑛 − 𝑎𝑛 𝑏0 ]
𝑥1
𝑥2
𝑥3
𝑦 = [1 0 0 … … … 0] ; + 𝑏0 𝑢
𝑥𝑛−1
[ 𝑥𝑛 ]

12 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Method III
Consider the following 𝑛𝑡ℎ order linear differential equation relating the output 𝑦(𝑡) to the input
𝑢(𝑡) of a system
Let 𝑛 = 𝑚 = 3
∴ 𝑦⃛ + 𝑎1 𝑦̈ + 𝑎2 𝑦̇ + 𝑎3 𝑦 = 𝑏0 𝑢
⃛ + 𝑏1 𝑢̈ + 𝑏2 𝑢̇ + 𝑏3 𝑢 (1)
On taking laplace transform of (1) with zero initial conditions,
𝑆 3 𝑌(𝑆) + 𝑎1 𝑆 2 𝑌(𝑆) + 𝑎2 𝑆𝑌(𝑆) + 𝑎3 𝑌(𝑆) = 𝑏0 𝑆 3 𝑈(𝑆) + 𝑏1 𝑆 2 𝑈(𝑆) + 𝑏2 𝑆𝑈(𝑆) + 𝑏3 𝑈(𝑆) (2)
𝑌(𝑆) 𝑏0 𝑆 3 +𝑏1 𝑆 2 +𝑏2 𝑆+𝑏3
𝑈(𝑆)
= 𝑆 3 +𝑎1 𝑆 2 +𝑎2 𝑆+𝑎3
𝑌(𝑆) 𝑋 (𝑆) 𝑌(𝑆)
Let = 1
𝑈(𝑆) 𝑈(𝑆) 𝑋1 (𝑆)
𝑋1 (𝑆) 1
Where, 𝑈(𝑆) = 𝑆 3 +𝑎 𝑆 2 +𝑎 𝑆+𝑎 and (3)
1 2 3
𝑌(𝑆) 3 2
= 𝑏0 𝑆 + 𝑏1 𝑆 + 𝑏2 𝑆 + 𝑏3 (4)
𝑋1 (𝑆)
On cross multiplying (3),
𝑋1 (𝑆)(𝑆 3 + 𝑎1 𝑆 2 + 𝑎2 𝑆 + 𝑎3 ) = 𝑈(𝑆)
𝑆 3 𝑋1 (𝑆) + 𝑎1 𝑆 2 𝑋1 (𝑆) + 𝑎2 𝑆𝑋1 (𝑆) + 𝑎3 𝑋1 (𝑆) = 𝑈(𝑆) (5)
On taking inverse laplace transform of (5),
𝑥⃛1 + 𝑎1 𝑥̈ 1 + 𝑎2 𝑥̇ 1 + 𝑎3 𝑥1 = 𝑢 (6)
Let the state variables be, 𝑥1 , 𝑥2 and 𝑥3
Where, 𝑥2 = 𝑥̇ 1 and
𝑥3 = 𝑥̈ 1 = 𝑥̇ 2 ; ∴ 𝑥̇ 3 = 𝑥⃛1 (7)
On substituting the state variables (7) in (6),
𝑥̇ 3 + 𝑎1 𝑥3 + 𝑎2 𝑥2 + 𝑎3 𝑥1 = 𝑢
𝑥̇ 3 = −𝑎3 𝑥1 − 𝑎2 𝑥2 − 𝑎1 𝑥3 + 𝑢
The state equations are,
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 3 = −𝑎1 𝑥3 − 𝑎2 𝑥2 − 𝑎3 𝑥1 + 𝑢
On cross multiplying equation (4),
𝑌(𝑆) = 𝑏0 𝑆 3 𝑋1 (𝑆) + 𝑏1 𝑆 2 𝑋1 (𝑆) + 𝑏2 𝑆𝑋1 (𝑆) + 𝑏3 𝑋1 (𝑆)
On taking inverse laplace transform,
𝑦 = 𝑏0 𝑥⃛1 + 𝑏1 𝑥̈ 1 + 𝑏2 𝑥̇ 1 + 𝑏3 𝑥1 (8)
On substituting the state variables (7) in (8),
𝑦 = 𝑏0 𝑥̇ 3 + 𝑏1 𝑥3 + 𝑏2 𝑥2 + 𝑏3 𝑥1 (9)
Put 𝑥̇ 3 = −𝑎3 𝑥1 − 𝑎2 𝑥2 − 𝑎1 𝑥3 + 𝑢
𝑦 = 𝑏0 (−𝑎3 𝑥1 − 𝑎2 𝑥2 − 𝑎1 𝑥3 + 𝑢) + 𝑏1 𝑥3 + 𝑏2 𝑥2 + 𝑏3 𝑥1
𝑦 = (𝑏3 − 𝑎3 𝑏0 )𝑥1 + (𝑏2 − 𝑎2 𝑏0 )𝑥2 + (𝑏1 −𝑎1 𝑏0 )𝑥3 + 𝑏0 𝑢 (10)
This equation is the output equation. On arranging the state equation and output equation in matrix
form,
𝑥̇ 1 0 1 0 𝑥1 0
𝑥̇
[ 2] = [ 0 0 1 𝑥
] [ 2 ] + [0] [𝑢] (11)
𝑥̇ 3 −𝑎3 −𝑎2 −𝑎1 𝑥3 1
𝑥1
𝑦 = [𝑏3 − 𝑎3 𝑏0 𝑏2 − 𝑎2 𝑏0 𝑏1 − 𝑎1 𝑏0 ] [𝑥2 ] + [𝑏0 ]𝑢 (12)
𝑥3
The above results can be generalized for an 𝑛𝑡ℎ order differential equation and the state model for
𝑚 = 𝑛 is given by,

13 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

𝑥̇ 1 0 1 0 0 ……… 0 𝑥1 0
𝑥̇ 2 0 0 1 0 ……… 0 𝑥2 0
𝑥̇ 3 0 0 0 1 ……… 0 𝑥3 0
;
=
; ; ; ; ; ; ; + ; [𝑢]
𝑥̇ 𝑛−1 0 0 0 0 ……… 1 𝑥𝑛−1 0
[ 𝑥̇ 𝑛 ] [−𝑎𝑛 −𝑎𝑛−1 −𝑎𝑛−2 −𝑎𝑛−3 … … … −𝑎1 ] [ 𝑥𝑛 ] [1]
𝑥1
𝑥2
𝑥
𝑦 = [(𝑏𝑛 − 𝑎𝑛 𝑏0 ) (𝑏𝑛−1 − 𝑎𝑛−1 𝑏0 ) … … … (𝑏2 − 𝑎2 𝑏0 ) (𝑏1 − 𝑎1 𝑏0 )] ;3 + 𝑏0 𝑢
𝑥𝑛−1
[ 𝑥𝑛 ]

Example 3: Obtain the state model of the system whose transfer function is given as,
𝒀(𝑺) 𝟏𝟎
= 𝟑
𝑼(𝑺) 𝑺 + 𝟒𝑺𝟐 + 𝟐𝑺 + 𝟏
Method I
𝑌(𝑆) 10
𝑈(𝑆)
= 𝑆 3 +4𝑆 2 +2𝑆+1 (1)
On cross multiplying (1),
𝑌(𝑆)(𝑆 3 + 4𝑆 2 + 2𝑆 + 1) = 10𝑈(𝑆)
𝑆 3 𝑌(𝑆) + 4𝑆 2 𝑌(𝑆) + 2𝑆𝑌(𝑆) + 𝑌(𝑆) = 10𝑈(𝑆) (2)
On taking inverse laplace transform of (2),
𝑦⃛ + 4𝑦̈ + 2𝑦̇ + 𝑦 = 10𝑢 (3)
Let the state variables be, 𝑥1 , 𝑥2 and 𝑥3
Where, 𝑥1 = 𝑦; 𝑥2 = 𝑦̇ ; ∴ 𝑥3 = 𝑦̈ (4)
On substituting the state variables (4) in (3),
𝑥̇ 3 + 4𝑥3 + 2𝑥2 + 𝑥1 = 10𝑢
𝑥̇ 3 = −𝑥1 − 2𝑥2 − 4𝑥3 + 10𝑢
The state equations are,
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 3 = −𝑥1 − 2𝑥2 − 4𝑥3 + 10𝑢 (5)
The output equation y= 𝑥1
The state model in the matrix form is,
𝑥̇ 1 0 1 0 𝑥1 0
[𝑥̇ 2 ] = [ 0 0 1 ] [𝑥2 ] + [ 0 ] [𝑢] (6)
𝑥̇ 3 −1 −2 −4 𝑥3 10
𝑥1
𝑦 = [1 0 0] [𝑥2 ] (7)
𝑥3
Method II
𝑌(𝑆) 10 10
𝑈(𝑆)
= 𝑆 3 +4𝑆 2 +2𝑆+1 = 4 2 1
𝑆 3 (1+ + 2 + 3 )
𝑆 𝑆 𝑆
10
𝑆3
= 4 2 1 (1)
1−(− − 2 − 3 )
𝑆 𝑆 𝑆
The signal flow graph for the above transfer function can be constructed with s single forward path
10
consisting of three integrators and with path gain 𝑆3
. The graph will have three individual loops
4 2 1
with loop gain − 𝑆 , − 𝑆 2 and − 𝑆 3 .

14 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Assign state variables at the output of the integrator. The state equations are obtained by summing
the incoming signals to the input of the integrators and equating them to the corresponding first
derivative of the state variable.
The state equations are,
𝑥̇ 1 = −4𝑥1 + 𝑥2
𝑥̇ 2 = −2𝑥1 + 𝑥3
𝑥̇ 3 = −𝑥1 + 10𝑢 (2)

The output equation is, 𝑦 = 𝑥1

The state model in the matrix form is,
𝑥̇ 1 −4 1 0 𝑥1 0
[𝑥̇ 2 ] = [−2 0 1] [𝑥2 ] + [ 0 ] [𝑢] (3)
𝑥̇ 3 −1 0 0 𝑥3 10
𝑥1
𝑦 = [1 0 0] [𝑥2 ] (4)
𝑥3

STATE SPACE REPRESENTATION USING CANONICAL VARIABLES

In canonical form the system matrix A will be a diagonal matrix. The elements are the poles of the
transfer function of the system.
Advantages: the state equations are independent of each other.
Disadvantages: canonical variables are not physical variables and so they are not available for
measurement and control.
By partial fraction expansion the transfer function 𝑛𝑡ℎ order system can be represented as
𝑌(𝑆) 𝐶1 𝐶2 𝐶
= 𝑏0 + + +, … … … … … + 𝑛 (1)
𝑈(𝑆) 𝑆+1 𝑆+2 𝑆+𝑛
Where 𝐶1 , 𝐶2 , 𝐶3 , … … . 𝐶𝑛 are the residues and 1 , 2 , 3 , … … . 𝑛 are roots of the denominator
polynomial (or poles of the system).

15 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

The equation (1) can be arranged as,

𝑌(𝑆) 𝐶1 𝐶2 𝐶𝑛
= 𝑏0 + 1 + 2 +, … … … … … + 
𝑈(𝑆) 𝑆(1+ ) 𝑆(1+ ) 𝑆(1+ 𝑛)
𝑆 𝑆 𝑆
𝐶1⁄ 𝐶2⁄ 𝐶𝑛⁄
𝑆 𝑆 𝑆
= 𝑏0 +  + 2 +, … … … … … + 
1+ 1 1+ 1+ 𝑛
𝑆 𝑆 𝑆
1⁄ 1⁄ 1⁄
𝑌(𝑆) = 𝑏0 𝑈(𝑆) + [ 1𝑆 × 𝐶1 ] 𝑈(𝑆) + [ 1
𝑆
× 𝐶2 ] 𝑈(𝑆)+, … … . + [ 1
𝑆
× 𝐶𝑛 ] 𝑈(𝑆) (2)
1+ ×1 1+ ×2 1+ ×𝑛
𝑆 𝑆 𝑆
This equation can be represented by a block diagram shown above

Assign state variables at the output of integrator. The input of the integrator will be first derivative
of state variable. The state equations are formed by adding the incoming signals and equating to
first derivative if state variable. The state equations are
𝑥̇ 1 = −1 𝑥1 + 𝑢
𝑥̇ 2 = −2 𝑥2 + 𝑢
𝑥̇ 3 = −3 𝑥3 + 𝑢
;
𝑥̇ 𝑛 = −𝑛 𝑥𝑛 + 𝑢
The output equation, 𝑦 = 𝐶1 𝑥1 + 𝐶2 𝑥2 +, … … . . +𝐶𝑛 𝑥𝑛 + 𝑏0 𝑢
The canonical form of state model in the matrix form is given by,
𝑥̇ 1 −1 0 0 0 ……… 0 𝑥1 1
𝑥̇ 2 0 −2 0 0 ……… 0 𝑥 2 1
𝑥̇ 3 0 0 −3 0 … … … 0 𝑥3 1
= ; + ; [𝑢] (3)
; ; ; ; ; ; ;
𝑥̇ 𝑛−1 0 0 0 0 ……… 0 𝑥𝑛−1 1
[ 𝑥̇ 𝑛 ] [ 0 0 0 0 … … … −𝑛 ] [ 𝑥𝑛 ] [1]
𝑥1
𝑥2
𝑥3
𝑦 = [𝐶1 𝐶2 𝐶3 … … … 𝐶𝑛 ] ; + 𝑏0 𝑢 (4)
𝑥𝑛−1
[ 𝑥𝑛 ]

When a pole of the transfer function has multiplicity, the canonical state model will be in a special
form called Jordan canonical form. In this form the system matrix
′𝐴′ will have a Jordan block of size 𝑞 × 𝑞, corresponding to a pole of value 1 , with multiplicity 𝑞. In
the Jordan block the diagonal element will be the poles and the element just above the diagonal is
one.
Consider a system with poles 1 , 1 , 1 , 4 , 5 , , … . 𝑛 , where 1 has multiplicity of three. The input
matrix (𝐵) and the system matrix for this case will be as shown in figure. The system matrix is also
denoted as 𝐽.
−1 1 0 0 ……… 0 0
0 −1 1 0 ……… 0 0
0 0 −1 1 ……… 0 1
𝐴=𝐽= and 𝐵 = (5)
0 0 0 −4 … … … 0 1
; ; ; ; ……… 1 ;
[ 0 0 0 0 … … … −𝑛 ] [1]
The transfer function of the system for this case in given by,
𝑌(𝑆) 𝐶 𝐶 𝐶 𝐶 𝐶
= 𝑏0 + (𝑆+1 3 + (𝑆+2 2 + (𝑆+3 ) + (𝑆+4 ) , … … … … … + 𝑆+𝑛 (6)
𝑈(𝑆) 1) 1) 1 4 𝑛
The block diagram of the system for this case is,
16 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Example 4: A feedback system has a closed loop transfer function

𝒀(𝑺) 𝟏𝟎(𝑺 + 𝟒)
=
𝑼(𝑺) 𝑺(𝑺 + 𝟏)(𝑺 + 𝟑)
Construct state models for this system using (i) Phase variable – method II (ii) Phase variable
– method III and (iii) Canonical variable. Give block diagram representation.

(i) Phase variable – method II

Y(S) 10(S+4) 10S+40 10S+40
U(S)
= S(S+1)(S+3) = S(S2 +4S+3) = S3 +4S2 +3S)
10 40
10𝑆+40 +
𝑆2 𝑆3
4 3 = 4 3 (1)
𝑆 3 (1+ + 2 ) 1−(− − 2 )
𝑆 𝑆 𝑆 𝑆

A signal flow graph for the above transfer function can be constructed with two forward paths and
10 40 4 3
two individual loops. The forward path gains are and . The loop gains are − and − 2 .
𝑆2 𝑆3 𝑆 𝑆

Assign state variables at the output of the integrator. The state equations are obtained by summing
the incoming signals to the input of the integrators and equating them to the corresponding first
derivative of the state variable.
The state equations are,
𝑥̇ 1 = −4𝑥1 + 𝑥2
𝑥̇ 2 = −3𝑥1 + 𝑥3 + 10𝑢
𝑥̇ 3 = 40𝑢 (2)
The output equation is, 𝑦 = 𝑥1
17 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

The state model is obtained by arranging the state equation and the output equations in the matrix
form.
𝑥̇ 1 −4 1 0 𝑥1 0
[𝑥̇ 2 ] = [−3 0 1] [𝑥2 ] + [10] [𝑢] (3)
𝑥̇ 3 0 0 0 𝑥3 40
𝑥1
[
𝑦 = 1 0 0 [ 2]] 𝑥 (4)
𝑥3
The block diagram representation of this state model is given as,

Using Phase variable – Method III

Y(S) 10(S+4)
U(S)
= S(S+1)(S+3)
𝑌(𝑆) 𝑋 (𝑆) 𝑌(𝑆) 10(S+4)
Let = 1 =
𝑈(𝑆) 𝑈(𝑆) 𝑋1 (𝑆) S(S+1)(S+3)
𝑋1 (𝑆) 1 𝑌(𝑆)
Where, 𝑈(𝑆) = S(S+1)(S+3) and 𝑋 (𝑆) = 10(S + 4)
1
𝑋1 (𝑆) 1 1 1
𝑈(𝑆)
= S(S+1)(S+3) = S(S2 +4S+3) = S3 +4S2 +3S (1)
On cross multiplying (1),
𝑋1 (𝑆)(𝑆 3 + 4𝑆 2 + 3𝑆) = 𝑈(𝑆)
𝑆 3 𝑋1 (𝑆) + 4𝑆 2 𝑋1 (𝑆) + 3𝑆𝑋1 (𝑆) = 𝑈(𝑆) (2)
On taking inverse laplace transform of (2),
𝑥⃛1 + 4𝑥̈ 1 + 3𝑥̇ 1 = 𝑢 (3)
Let the state variables be, 𝑥1 , 𝑥2 and 𝑥3
Where, 𝑥2 = 𝑥̇ 1 and
𝑥3 = 𝑥̈ 1 = 𝑥̇ 2 ; ∴ 𝑥̇ 3 = 𝑥⃛1 (4)
On substituting the state variables (4) in (3),
𝑥̇ 3 + 4𝑥3 + 3𝑥2 = 𝑢
𝑥̇ 3 = −3𝑥2 − 4𝑥3 + 𝑢
The state equations are,
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 3 = −3𝑥2 − 4𝑥3 + 𝑢 (5)
Consider the second part of transfer function,
𝑌(𝑆)
𝑋1 (𝑆)
= 10(S + 4) = 10𝑆 + 40 (6)
On cross multiplying equation (6),
𝑌(𝑆) = 10𝑆𝑋1 (𝑆) + 40𝑋1 (𝑆)
On taking inverse laplace transform,
𝑦 = 10𝑥̇ 1 + 40𝑥1
Put 𝑥̇ 1 = 𝑥2
𝑦 = 10𝑥2 + 40𝑥1 = 40𝑥1 + 10𝑥2
This equation is the output equation. On arranging the state equation and output equation in matrix
form,
18 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

𝑥̇ 1 10 0 𝑥1 0
[𝑥̇ 2 ] = [0
0 1 ] [𝑥2 ] + [0] [𝑢] (7)
𝑥̇ 3 −3 −4 𝑥3
0 1
𝑥1
𝑦 = [40 10 0] [𝑥2 ] (8)
𝑥3
The block diagram representation of this state model is,

(iii) Using Canonical variable

Y(S) 10(S + 4)
=
U(S) S(S + 1)(S + 3)
Y(S)
By partial fraction expansion the transfer function U(S) can be represented as
𝑌(𝑆) 10(S+4) 𝐴 𝐵 𝐶
= = + + (1)
𝑈(𝑆) S(S+1)(S+3) 𝑆 𝑆+1 𝑆+3

10(S+4) 10×4 40
𝐴 = (S+1)(S+3) | = =
𝑆=0 1×3 3

10(S+4) 10(−1+4) 10×3

𝐵= |
𝑆(S+3) 𝑆=−1
= −1(−1+3) = −1×2 = −15

10(S+4) 10(−3+4) 10×1 5

𝐶= |
𝑆(S+1) 𝑆=−3
= −1(−3+1) = −3×(−2) = 3

𝑌(𝑆) 40⁄ 15 5⁄
3 3
𝑈(𝑆)
= 𝑆
− 𝑆+1 + 𝑆+3 (2)

The equation (2) can be rearranged as,

𝑌(𝑆) 40⁄ 15 5⁄
3 3
𝑈(𝑆)
= 𝑆
− 1 + 3
𝑆(1+ ) 𝑆(1+ )
𝑆 𝑆

1 40 1⁄ 1⁄ 5
𝑆 𝑆
𝑌(𝑆) = [𝑆 × 3
] 𝑈(𝑆) − [ 1 × 15] 𝑈(𝑆) + [ 1 × 3] 𝑈(𝑆) (3)
1+ ×1 1+ ×3
𝑆 𝑆
This equation can be represented by a block diagram

19 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Assign state variables at the output of integrator. The input of the integrator will be first derivative
of state variable. The state equations are formed by adding the incoming signals and equating to
first derivative if state variable. The state equations are
𝑥̇ 1 = 𝑢
𝑥̇ 2 = −𝑥2 + 𝑢
𝑥̇ 3 = −3𝑥3 + 𝑢
40 5
The output equation, 𝑦 = 3 1
𝑥 − 15𝑥2 + 3 𝑥3
The canonical form of state model in the matrix form is given by,
𝑥̇ 1 0 0 0 𝑥1 1
[𝑥̇ 2 ] = [0 −1 0 ] [𝑥2 ] + [1] [𝑢] (4)
𝑥̇ 3 0 0 −3 𝑥3 1
𝑥1
40 5
𝑦 = [ 3 −15 3] [𝑥2 ] (5)
𝑥3

SOLUTIONS OF STATE EQUATION

Solution of homogeneous state equation(without input/excitation)
Consider the state equation without input vector, 𝑋̇ = 𝐴𝑋(𝑡)
On taking laplace transform,
𝑠𝑋(𝑠) − 𝑋(0) = 𝐴𝑋(𝑠)
𝑠𝑋(𝑠) − 𝐴𝑋(𝑠) = 𝑋(0)
𝑠[𝑋(𝑠) − 𝐴𝑋(𝑠)] = 𝑋(0)
[𝑠𝐼 − 𝐴]𝑋(𝑠) = 𝑋(0)
Premultiply both sides by [𝑠𝐼 − 𝐴]−1
𝑋(𝑠) = [𝑠𝐼 − 𝐴]−1 𝑋(0)
On taking inverse laplace transform,
𝑋(𝑡) = 𝐿−1 [[𝑠𝐼 − 𝐴]−1 𝑋(0)] = 𝐿−1 [[𝑠𝐼 − 𝐴]−1 ] 𝑋(0) (1)
𝐴𝑡
On comparing equation (1) with the solution to state equation, 𝑋(𝑡) = 𝑒 𝑋(0)
𝑒 𝐴𝑡 = 𝐿−1 [[𝑠𝐼 − 𝐴]−1 ] or 𝐿[𝑒 𝐴𝑡 ] = [𝑠𝐼 − 𝐴]−1 (2)
𝐴𝑡 𝐴𝑡 ]
𝑒 = ∅(𝑡); 𝐿[𝑒 = 𝐿[∅(𝑡)] = ∅(𝑠) (3)
−1
Where, ∅(𝑠) = [𝑠𝐼 − 𝐴] and it is called resolvant matrix.
From the system matrix 𝐴, the resolvant matrix ∅(𝑠) can be computed. By taking inverse laplace
transform of resolvant matrix the state transition matrix is computed from which the solution of
state equation is obtained. The solution of state equation is given by,
𝑋(𝑡) = 𝑒 𝐴𝑡 𝑋(0)
𝑋(𝑡) = 𝐿−1 [∅(𝑠)] 𝑋(0)
Where ∅(𝑠) = [𝑠𝐼 − 𝐴]−1
Solution of non homogeneous state equation(with input/excitation)
The state equation of 𝑛𝑡ℎ order system is give by,
𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡); 𝑋(0) = 𝑋0 (1)
Where 𝑋(0) is initial condition vector.
On taking laplace transform of equation (1)
𝑠𝑋(𝑠) − 𝑋(0) = 𝐴𝑋(𝑠) + 𝐵𝑈(𝑠)
𝑠𝐼𝑋(𝑠) − 𝐴𝑋(𝑠) = 𝑋(0) + 𝐵𝑈(𝑠), where 𝐼 is the unit matrix
[𝑠𝐼 − 𝐴]𝑋(𝑠) = 𝑋(0) + 𝐵𝑈(𝑠) (2)
−1
Premultiply both sides by [𝑠𝐼 − 𝐴]
𝑋(𝑠) = [𝑠𝐼 − 𝐴]−1 𝑋(0) + [𝑠𝐼 − 𝐴]−1 𝐵𝑈(𝑠)
20 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

= ∅(𝑠)𝑋(0) + ∅(𝑠)𝐵𝑈(𝑠) (3)

On taking inverse laplace transform,
𝑋(𝑡) = 𝐿−1 [∅(𝑠)𝑋(0)] + 𝐿−1 [∅(𝑠)𝐵𝑈(𝑠)]
𝑋(𝑡) = ∅(𝑡)𝑋(0) + 𝐿−1 [∅(𝑠)𝐵𝑈(𝑠)] (4)
The equation (4) is the solution of state equation with input/forcing function.

Example 5: A linear time invariant system is characterized by homogeneous state equation

ẋ 1 0 x1
[ 1] = [ ][ ]
ẋ 2 1 1 x2
1
Compute the solution of the homogeneous equation, assuming the initial state vector x0 = [ ]
0
Solution

A linear time invariant system is characterized by homogeneous state equation

ẋ 1 0 x1
[ 1] = [ ][ ]
ẋ 2 1 1 x2
1
Compute the solution of the homogeneous equation, assuming the initial state vector x0 = [ ]
0

linear time invariant system is characterized by homogeneous state equation

ẋ 1 0 x1
[ 1] = [ ][ ]
ẋ 2 1 1 x2
Compute the solution of the homogeneous equation, assuming the initial state vector

1 0 1
Given, 𝐴 = [ ] ; 𝑥0 = [ ]
1 1 0
Solution of the state equation
𝑋(𝑡) = 𝑒 𝐴𝑡 𝑋(0); 𝑒 𝐴𝑡 = 𝐿−1 [[𝑠𝐼 − 𝐴]−1 ]

[𝑠𝐼 − 𝐴] = 𝑠 [1 0] − [1 0
]=[
𝑠 0
]−[
1 0
]=[
𝑠−1 0
]
0 1 1 1 0 𝑠 1 1 −1 𝑠−1

𝑠−1 0
| | = (𝑠 − 1)2 − 0 = (𝑠 − 1)2
−1 𝑠 − 1
1
1 0
[𝑠 − 1]−1 = 𝑠−1 0 𝑠−1
[ ] = 1 1
(𝑠 − 1)2 −1 𝑠 − 1
2
[(𝑠 − 1) 𝑠 − 1]
−𝑡
𝑒 𝐴𝑡 = ∅(𝑡) = 𝐿−1 [∅(𝑠)] = 𝐿−1 [𝑠𝐼 − 𝐴]−1 = [ 𝑒 −𝑡 0−𝑡 ]
𝑡𝑒 𝑒
−𝑡
−1 𝐴𝑡
The solution to state equation 𝑋(𝑡) = 𝐿 𝑒 𝑋(0) = [ −𝑡 𝑒 0 ] × [1] = [ 𝑒 −𝑡 ]
𝑡𝑒 𝑒 −𝑡 0 𝑡𝑒 −𝑡
σ 0 0 ω σ ω
Example 6: given that A1 = [ ] ; A2 = [ ];A = [ ] compute eAt
0 σ −ω 0 −ω σ

𝜎 0 0 𝜔 𝜎 𝜔
given that 𝐴1 = [ ] ; 𝐴2 = [ ];𝐴 = [ ] 𝑐𝑜𝑚𝑝𝑢𝑡𝑒 𝑒 𝐴𝑡
0 𝜎 −𝜔 0 −𝜔 𝜎

21 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Solution
Here, 𝐴 = 𝐴1 + 𝐴2
𝑒 𝐴𝑡 = 𝑒 (𝐴1 +𝐴2 )𝑡 = 𝑒 𝐴1 𝑡 . 𝑒 𝐴2 𝑡
[𝑠𝐼 − 𝐴1 ] = 𝑠 [1 0] − [𝜎 0] = [ 𝑠 0
]−[
𝜎 0
]=[
𝑠−𝜎 0
]
0 1 0 𝜎 0 𝑠 0 𝜎 0 𝑠−𝜎

|𝑠𝐼 − 𝐴1 | = |𝑠 − 𝜎 0
| = (𝑠 − 𝜎)2 − 0 = (𝑠 − 𝜎)2
0 𝑠−𝜎
1
1 0
[𝑠 − 𝐴1 ]−1 = 𝑠−𝜎 0 𝑠−𝜎
[ ] = [ ]
(𝑠 − 𝜎)2 0 𝑠−𝜎 1
0
𝑠−𝜎
−𝜎𝑡
𝑒 𝐴1 𝑡
=𝐿−1 [𝑠𝐼 −1
− 𝐴1 ] = [ 𝑒 0 ]
0 𝑒 −𝜎𝑡

𝑠 −𝜔
[𝑠𝐼 − 𝐴2 ] = 𝑠 [1 0
]−[
0 𝜔
]=[
𝑠 0
]−[
0 𝜔
]=[ ]
0 1 −𝜔 0 0 𝑠 −𝜔 0 𝜔 𝑠

𝑠 −𝜔
|𝑠𝐼 − 𝐴2 | = | | = 𝑠 2 + 𝜔2
𝜔 𝑠
𝑠2 𝜔
1 𝑠 −𝜔 2 2 𝑠2 + 𝜔2
[𝑠 − 𝐴2 ]−1 = 2 [ ]= 𝑠 +𝜔
𝑠 +𝜔 2 𝜔 𝑠 −𝜔 𝑠2
[𝑠 2 + 𝜔 2 𝑠2 + 𝜔2]

cos 𝜔𝑡 sin 𝜔𝑡
𝑒 𝐴2 𝑡 = 𝐿−1 [𝑠𝐼 − 𝐴2 ]−1 = [ ]
− sin 𝜔𝑡 cos 𝜔𝑡
−𝜎𝑡
𝑒 𝐴𝑡 = 𝑒 𝐴1 𝑡 . 𝑒 𝐴2 𝑡 = [𝑒 0 ] [ cos 𝜔𝑡 sin 𝜔𝑡
]
0 𝑒 −𝜎𝑡 − sin 𝜔𝑡 cos 𝜔𝑡
−𝜎𝑡
𝑒 𝐴𝑡 = [ 𝑒 −𝜎𝑡cos 𝜔𝑡 𝑒 −𝜎𝑡 sin 𝜔𝑡 ]
−𝑒 sin 𝜔𝑡 𝑒 −𝜎𝑡 cos 𝜔𝑡

CONCEPTS OF CONTROLLABILITY & OBSERVABILITY

Controllability: the controllability verifies the usefulness of a state variable. In the controllability
test we can find whether the state variable can be controlled to achieve the desired output. The choice
of state variable is arbitrary while forming the state model. After determining the state model, the
controllability of the state variable is verified. If the state variable is not controllable then we go for
another choice of state variable.

Definition: A system is said to be completely state controllable if it is possible to transfer the

system state from any initial state 𝑿(𝒕𝟎 ) any desired state 𝑿(𝒕) in specified finite time by a control
vector 𝑼(𝒕).
Kalman‟s test: Consider a system with state equation 𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡). For this system, a
composite matrix 𝑄𝐶 can be formed such that,
𝑄𝐶 = [𝐵 𝐴𝐵 𝐴2 𝐵 , … … . 𝐴𝑛−1 𝐵]
Where 𝑛 is the order of the system( 𝑛 is also equal to number of state variables).
The system is completely state controllable if the rank of the composite matrix 𝑄𝐶 is 𝑛.
𝑖. 𝑒. |𝑄𝐶 | = |𝐵 𝐴𝐵 𝐴2 𝐵 , … … . 𝐴𝑛−1 𝐵| ≠ 0
Advantages: calculations are simple.
Disadvantage: we cannot find the state variable which is uncontrollable.
22 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Observability: here, we can find whether the state variable is observable or measurable. The
concept of observability is useful in solving the problem of reconstructing unmeasurable state
variables from measurable ones in the minimum possible length of time.

Definition: A system is said to be completely observable, if every state 𝑿(𝒕) can be completely
identified by measurements of the outputs 𝒀(𝒕) over a finite time interval.

Kalman‟s test: Consider a state model of 𝑛𝑡ℎ system 𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡); 𝑌 = 𝐶𝑋 + 𝐷𝑈. For this
system, a composite matrix, 𝑄𝐶 can be formed such that,
𝑄𝑜 = [𝐶 𝑇 𝐴𝑇 𝐶 𝑇 (𝐴𝑇 )2 𝐶 𝑇 (𝐴𝑇 )3 𝐶 𝑇 … … … . . (𝐴𝑇 )𝑛−1 𝐶 𝑇 ]
Where 𝑛 is the order of the system( 𝑛 is also equal to number of state variables).
The system is completely observable if the rank of the composite matrix 𝑄𝐶 is 𝑛.
𝒊. 𝒆. , |𝑸𝑪 | = |𝐶𝑇 𝐴𝑇 𝐶𝑇 (𝐴𝑇 )2 𝐶𝑇 (𝐴𝑇 )3 𝐶𝑇 ……….. (𝐴𝑇 )𝑛−1 𝐶𝑇 | ≠ 𝟎

Relationship between construability and observability-Duality Principle

The concept of controllability and observability play an important role in the design of control
system in state space. They govern the existence of a complete solution to the control system design
problem. The solution to this problem may not exists if the system considered is not controllable. It
is important to note that all physical systems are controllable and observable. However the
mathematical model of these systems may not possess the property of the controllability or
observability. Then it is necessary to know the condition under which a system is controllable and
observable and the designer can seek another state model which is controllable and observable.
Duality principle.
The principle duality states that a system is completely state controllable if and only if its dual system
is completely observable and vice versa.(if the system is observable then its dual is controllable).
Using this principle of duality, the observability of a given system can be checked by testing the state
controllability of its dual or vice versa.
Consider the system 𝑆1 described the state model,
𝑋̇(𝑡) = 𝐴𝑋(𝑡) + 𝐵𝑈(𝑡)
𝑌 = 𝐶𝑋
Let the dual of system 𝑆1 be denoted as 𝑆2 and the dual system 𝑆2 is described the state model,
𝑍̇ = 𝐴𝑇 𝑍 + 𝐶 𝑇 𝑉
𝑁 = 𝐵𝑇 𝑍
Where, 𝑍 = State vector of dual system, 𝑉 = Input vector of dual system; 𝑁 = Output vector of dual
system.
For the system 𝑆1 the composite matrix 𝑄𝑐1 for controllability id given by equation (1) and the
composite matrix, 𝑄𝑐2 for observability is given by equation (2),
𝑄𝐶1 = [𝐵 𝐴𝐵 𝐴2 𝐵 , … … . 𝐴𝑛−1 𝐵] (1)
𝑄𝑜1 = [𝐶 𝑇 𝑇
𝐴 𝐶 𝑇 𝑇 2
(𝐴 ) 𝐶 𝑇 𝑇 3
(𝐴 ) 𝐶 𝑇 𝑇 𝑛−1
… … … . . (𝐴 ) 𝐶 ]𝑇 (2)
For the dual system 𝑆2 the composite matrix, 𝑄𝐶2 for controllability is given by equation (3) and
composite matrix 𝑄𝑜2 for observability is given by equation (4)
𝑄𝐶2 = [𝐶 𝑇 𝐴𝑇 𝐶 𝑇 (𝐴𝑇 )2 𝐶 𝑇 (𝐴𝑇 )3 𝐶 𝑇 … … … . . (𝐴𝑇 )𝑛−1 𝐶 𝑇 ] (3)
2
𝑄𝑂2 = [𝐵 𝐴𝐵 𝐴 𝐵 , … … . 𝐴 𝐵] 𝑛−1 (4)
From equations (1) and (4), 𝑄𝐶1 = 𝑄𝑂2 , hence if the system 𝑆1 is controllable then its dual system 𝑆2
is observable. From equations (2) and (3), 𝑄𝑂1 = 𝑄𝐶2 , hence if the system 𝑆1 is observable then its
dual system 𝑆2 is controllable.

23 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

Example 7: Write the state equations for the system shown in which 𝑥1 , 𝑥2 and 𝑥3 constitute the state
vector. Determine whether the system is completely controllable and observable.

The state equations are obtained by writing equations for the output of each block and then taking
inverse laplace. With reference to figure 1,
2
𝑋1 (𝑆) = [𝑋2 (𝑆) − 𝑋3 (𝑆)] [ ]
𝑆(𝑆 + 1)
𝑆(𝑆 + 1)𝑋1 (𝑆) = 2𝑋2 (𝑆) − 2𝑋3 (𝑆)
𝑆 2 𝑋2 (𝑆) + 𝑆𝑋1 (𝑆) = 2𝑋2 (𝑆) − 2𝑋3 (𝑆)
On taking inverse laplace transform,
𝑥̈ 1 + 𝑥̇ 1 = 2𝑥2 − 2𝑥3 (1)
With reference to figure 2,
𝑋3 (𝑆) = 𝑆𝑋1 (𝑆)
On taking inverse laplace transform,
𝑥3 = 𝑥̇ 1 (2)
With reference to figure 3,
2
𝑋2 (𝑆) = [𝑈(𝑆) − 𝑋1 (𝑆)] [ ]
𝑆+3
(𝑆 + 3)𝑋2 (𝑆) = 2𝑈(𝑆) − 2𝑋1 (𝑆)
𝑆𝑋2 (𝑆) + 3𝑋2 (𝑆) = 2𝑈(𝑆) − 2𝑋1 (𝑆)
On taking inverse laplace transform,
𝑥̇ 2 + 3𝑥2 = 2𝑢 − 2𝑥1
𝑥̇ 2 = −2𝑥1 − 3𝑥2 + 2𝑢 (3)
From equation 2,
𝑥̇ 1 = 𝑥3 ; ∴ 𝑥̈ 1 = 𝑥̇ 3
Put 𝑥̇ 1 = 𝑥3 and 𝑥̈ 1 = 𝑥̇ 3 in equation (1),
𝑥̇ 3 + 𝑥3 = 2𝑥2 − 2𝑥3
𝑥̇ 3 = 2𝑥2 − 2𝑥3 − 𝑥3
𝑥̇ 3 = 2𝑥2 − 3𝑥3 (4)
The state equations are given by,
𝑥̇ 1 = 𝑥3
𝑥̇ 2 = −2𝑥1 − 3𝑥2 + 2𝑢
𝑥̇ 3 = 2𝑥2 − 3𝑥3
The output equation is y = 𝑥1
The state model in the matrix form
𝑥̇ 1 0 0 1 𝑥1 0
𝑥̇
[ 2 ] [−2 −3 0 ] [ 2 ] [2] [𝑢]
= 𝑥 +
𝑥̇ 3 0 2 −3 𝑥3 0
𝑥1
𝑦 = [1 0 0] [𝑥2 ]
𝑥3
Kalman test for controllability
|𝑄𝐶 | = |𝐵 𝐴𝐵 𝐴2 𝐵 , … … . 𝐴𝑛−1 𝐵| ≠ 0
24 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

0 0 1 0
𝐴 = [−2 −3 0 ] ; 𝐵 = [2]
0 2 −3 0
0 0 1 0 0
𝐴 × 𝐵 = [−2 −3 0 ] × [2] = [−6]
0 2 −3 0 0
0 0 1 0 0 1 0 2 −3
𝐴2 = 𝐴 × 𝐴 = [−2 −3 0 ] × [−2 −3 0 ] = [ 6 9 −2]
0 2 −3 0 2 −3 −4 −12 9
0 2 −3 0 4
𝐴2 𝐵 = 𝐴2 × 𝐵 = [ 6 9 −2] × [2] = [ 18 ]
−4 −12 9 0 −24
0 0 4
The composite matrix for controllability 𝐶 = 𝐵 𝐴𝐵 𝐴 𝐵 = [2 −6
[𝑄 ] [ 2 ] 18 ]
0 4 −24
0 0 4
Determinant of 𝑄𝐶 = |2 −6 18 | = 0 − 0 + 4(8 − 0) = 32 ≠ 0
0 4 −24
The rank of 𝑄𝐶 = 3, hence the system is completely state controllable.
Kalman test for observability
0 0 1
𝐴 = [−2 −3 0 ] ; 𝐶 = [1 0 0]
0 2 −3
1
𝐶 𝑇 = [1 0 0]𝑇 = [0]
0
0 0 1 𝑇 0 −2 0
𝐴𝑇 = [−2 −3 0 ] = [0 −3 2 ]
0 2 −3 1 0 −3
0 −2 0 0 −2 0 0 6 −4
(𝐴𝑇 )2 = [0 −3 2 ] × [0 −3 2 ] = [ 2 9 −12]
1 0 −3 1 0 −3 −3 −2 9
0 −2 0 1 0
𝐴𝑇 𝐶 𝑇 = [0 −3 2 ] × [0] = [0]
1 0 −3 0 1
0 6 −4 1 0
(𝐴𝑇 )2 𝐶 𝑇 = [ 2 9 −12 ] × [0 ] = [ 2]
−3 −2 9 0 −3
1 0 0
The composite matrix for observability, 𝑄𝑜 = [𝐶 𝑇 𝐴𝑇 𝐶 𝑇 (𝐴𝑇 )2 𝐶 𝑇 ] = [0 0 2]
0 1 −3
1 0 0
Determinant of , 𝑄𝑜 = |0 0 2 | = 1 × −2 = −2
0 1 −3
Since , |𝑄𝑜 | ≠ 0, the rank of 𝑄𝑜 = 3
Hence the system is completely observable.

Prepared By,

******************************************* ALL THE BEST**************************************

25 VI Semester ICE
IC8651 Advanced Control System Department of ICE 2022-23

26 VI Semester ICE