Sem.

2
Preliminaries to State-Space Methods 2019/2020

 State variable analysis is a time domain approach, but has several advantages over
classical control methods.

lmportant Definitions
 State: The state of a dynamic system is defined as n minimal set of variables such
that the knowledge of these variables at together with the knowledge of the
Inputs for , completely determines the behavior of the system for .
 State Variables: The variables involved in determining the state of a dynamic
system X(t), are called the state variables. X 1 (t) , X2(t) ...... Xn(t) are nothing but
the state variables. These are normally the energy storing elements contained in
the system.
 State Space: The space whose co‐ordinate axes are nothing but the ‘n’ state variables with 
time as the implicit variable is called the state space.
 State Trajectory: It is the locus of the tips of the state vectors, with time as the implicit 
variable.

Some of the advantages of state space approach

 The method takes into account the effect of all initial conditions.
 It can be applied to nonlinear as well as time varying systems.
 It can be conveniently applied to Multiple Input Multiple Output systems.
 The system can be designed for the optimal conditions precisely by using this
modern method.

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 1

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

 As the method involves matrix algebra, can be conveniently adopted for the digital
computers.
 The state variables selected need not necessarily be the physical quantities of the
system.

State model of a linear system

 Consider a multi-input multi-output (MIMO) system of an nth order illustrated in
the figure below

 Number of inputs=m, while number ofoutputs=p

 In state variable representation, such a system can be arranged in the form of ‘n‘
first order differential equation as follows

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 2

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

where is a functional operator.

 The functional output is given by , where the functional operator.
 For time variant system, the dynamic system is represented as

 For linear time invariant system, state space representation is given by

where
: State vector matrix of order nx1
: Input vector matrix of order mx1
: Output vector matrix of order px1
: System matrix of order nxn
: Input matrix of order nxm
: Output matrix of order pxn, : Direct transmission matrix of order pxm
Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 3
 Sem. 2
Preliminaries to State-Space Methods 2019/2020

State diagram for a standard state model

Consider a standard state-space model given by

The corresponding state diagram is given as

General remarks about state diagram

 To obtain the state model from state diagram, always choose output of each
Integrator as a stale variable. Number of integrators always equals the order of the
system
 Differentiators arc not used in the state diagram as they amplify the inevitablc noise

State variable representation using physical variables

Generally, physical variables associated with energy storing elements, which are
responsible for initial conditions, are selected as state variable of a given system.
Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 4
 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Example: Electrical System

Obtain the state model of the following electrical system

Solution:
 There are two energy storing elements, L and C, so the two state variables are
current through inductor and voltage across the capacitor,
i.e and and is input variable
Applying KVL to the loop

Rearranging the equation

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 5

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Voltage across the capacitor Example: Mechanical System

Obtain the state model of the following
mechanical system

Therefore,

From which

i.e.,

Output variable is given by

 System equation (2nd order system)

Implying
 Defining state variable as

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 6

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

NB: The state model is not unique

 Therefore property of the system.

Advantages of using physical

variables as state variables
or  Physical quantities selected as state
variables can be measured.
 It is easier to design a state feedback
since the feedback may consist of state
variables as well as outputs.
 The output equation  The behavor of various physical
variables is available with time.

 In a vector-matrix form State space representation using

Canonical variables
 In canonical variable representation,
matrix A is a diagonal matrix and plays
an important role in state-space
analysis.
 The output

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 7

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Linear Time-Invariant (LTI) System

 A dynamic system is called linear if the following conditions are fulfilled
– Superposition principle

–Principle of amplification

 A dynamic system is called time-invariant, if for any function , the equation

is valid for all , where the functions and
are introduced.

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 8

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Transfer function from a State Model  Hence the transfer function is given by
 Consider a stand state model of a LTI
system

 Where,

 Taking Laplace transform,

 Characteristic equation

 Zero initial conditons are assumed  The stability of the system depends on
for transfer function, hence the roots of characteristic equation
(eigen values of matrix

 Premultiplyig both sides by

 Substituting into the output equation

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 9

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Solution of the State Equation  Then, the solution of the state

 Consider the following State equation differential equation is given by

with as initial conditions

 The expenential matrix is called
Nonhomogeneous solution state transition matrix and it
 Taking the Laplace transform of the represents the natural response of the
state equation system

 Therefore,  And

 Inverse Laplace transformation  Hence,

 The matrix exponential function is

defined as

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 10

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Controllable Canonical Form

 Consider an nth order linear plant model described by the differential equation

 Where y(t) is the plant output and u(t) is the plant input
 The state variables (phase variables are defines as

 Taking derivatives of the first n-1 state variables

 In matrix form

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 11

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Output equation

Control Canonical Form (Example)

 Consider the transfer function of a third-order

 Decomposed transfer function in a cascaded form

𝑈 𝑠 1 𝑊 𝑠
𝑌 𝑠
𝑏 𝑠 𝑏 𝑠 𝑏
𝑠 𝑎 𝑠 𝑎 𝑠 𝑎

 From which

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 12

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

 Choosing the state variable in phase variable form

 The output equation is given by

 In vector matrix form

Transformation to Controllable Canonical Form (CCF)

 Transformation to CCF is done by means of transformation matrix

 Where is controllable matrix given by

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 13

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

 The denotes coefficient matrix given by

 Here, are coefficients of the characteristi polynomial

 The transformed matrices are given as

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 14

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Observable Canonical Form

 Consider an nth order linear plant model described by the differential equation

 The following state-space representation is referred to as an observable canonical

form

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 15

 Sem. 2
Preliminaries to State-Space Methods 2019/2020

Transformation to Observable Canonical Form (OCF)

 Transformation is done using the following transformation Matric

 Where is observability matrix given as

 The denotes coefficient matrix given by

 Here, are coefficients of the characteristi polynomial

 The transformed matrices are given as

Dr.-Ing. Jackson G. Njiri EMT 3104 L-1: Preliminaries to State-Space Methods 16