Namaste

Control System

Shahabuddin Khan
Assistant Professor
Department of Electrical Engineering
Pulchowk Campus, IOE
[email protected]
 Chapter – 8
State Space Analysis [4 Hrs.]
Shahabuddin Khan
Assistant Professor
Department of Electrical Engineering
Pulchowk Campus, IOE
[email protected]
 Syllabus
8.1. Definition of state -space
8.2. State space representation of electrical and mechanical system
8.3. Conversion from state space to transfer function
8.4. Conversion from transfer function to state space
8.5. State-transition matrix

7/10/22 10:52 PM SK, Pulchowk Campus 4

 Syllabus
8.1. Definition of state -space
8.2. State space representation of electrical and mechanical system
8.3. Conversion from state space to transfer function
8.4. Conversion from transfer function to state space
8.5. State-transition matrix

7/10/22 10:52 PM SK, Pulchowk Campus 5

 Introduction
Control System Design
and Analysis

Modern
Conventional
Approach
Approach
[Time Series
[Root Locus,
and State
Bode Plot,
Space
Nyquist Plot]
Analysis]

7/10/22 10:52 PM SK, Pulchowk Campus 6

 Limitation of Transfer Function Approach
1. Naturally, significant initial conditions in obtaining precise solution of any
system, loose their importance in conventional approach.
2. The method is insufficient and troublesome to give complete time domain
solution of higher order systems.
3. It is not very much convenient for the analysis of Multiple Input Multiple Output
systems.
4. It gives analysis of system for some specific types of inputs like Step, Ramp
etc.
5. It is only applicable to Linear Time Invariant Systems.
6. The classical methods like Root locus, Bode plot etc are basically trial and error
procedures which fail to give the optimal solution required.

7/10/22 10:52 PM SK, Pulchowk Campus 7

 State Space Analysis
v The technique which uses the concept of state is called state space analysis or
state variable analysis.
v It is a modern method also a time domain approach which uses the concept of
total internal state of the system considering all initial conditions.

7/10/22 10:52 PM SK, Pulchowk Campus 8

 Concept of State

7/10/22 10:52 PM SK, Pulchowk Campus 9

 Concept of State

State can be represented

as a compact and concise
representation of the past
history of the system

7/10/22 10:52 PM SK, Pulchowk Campus 10

 Dynamic and Static System
System in which the output
is not only dependent on
the input but also
dependent on the initial
conditions are called the
system with memory or
dynamic systems.
System in which the output
of the system is only
dependent on the input
applied at t=0 are called the
system with zero memory
or static systems.

7/10/22 10:52 PM SK, Pulchowk Campus 11

 Critical State

State which affect the system desired output are regarded as critical state.

7/10/22 10:52 PM SK, Pulchowk Campus 12

 Advantages of State Space Analysis
1. The method takes into account the effect of all initial conditions.
2. It can be applied to nonlinear as well as time varying systems.
3. It can be conveniently applied to Multiple Input Multiple Output systems.
4. The system can be designed for the optimal conditions precisely by using this
modern method.
5. Any type of the input can be considered for designing the system.
6. As the method involves matrix algebra, can be conveniently adopted for the
digital computers.
7. The state variables selected need not necessarily be the physical quantities of
the system.
8. The vector matrix notation greatly simplifies the mathematical representation
of the system.

7/10/22 10:52 PM SK, Pulchowk Campus 13

 Important Definitions
State :
The state of a dynamic system is defined as a minimal set of variables such
that the knowledge of these variables at 𝑡 = 𝑡! together with the knowledge of the
inputs for 𝑡 ≥ 𝑡", completely determines the behavior of the system for 𝑡 > 𝑡".
State Variables :
The variables involved in determining the state of a dynamic system X(t),
are called the state variables. 𝑋# 𝑡 , 𝑋$ 𝑡 … … . . 𝑋% (𝑡) are nothing but the state
variables. These are normally the energy storing elements contained in the system.
State Vector :
The 'n' state variables necessary to describe the complete behavior of the
system can be considered as 'n' components of a vector X(t) called the state
vector at time ‘𝑡’. The state vector X(t) is the vector sum of all the state variables.

7/10/22 10:52 PM SK, Pulchowk Campus 14

 Important Definitions
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.

7/10/22 10:52 PM SK, Pulchowk Campus 15

 Important Definitions: Example
Lets take a second order system.
Since Order is two, We will have two states.
Lets say 𝑋#(𝑡) and 𝑋$(𝑡)à State Variables
𝑋# 𝑡
𝑋 𝑡 = ⇒ 𝑆𝑡𝑎𝑡𝑒 𝑉𝑒𝑐𝑡𝑜𝑟
𝑋$ 𝑡
𝑋$ 𝑡 𝑋 𝑡$
𝑋 𝑡#
𝑆𝑡𝑎𝑡𝑒 𝑇𝑟𝑎𝑗𝑒𝑐𝑡𝑜𝑟𝑦

𝑋# 𝑡

7/10/22 10:52 PM SK, Pulchowk Campus 16

 Syllabus
8.1. Definition of state -space
8.2. State space representation of electrical and mechanical system
8.3. Conversion from state space to transfer function
8.4. Conversion from transfer function to state space
8.5. State-transition matrix

7/10/22 10:52 PM SK, Pulchowk Campus 17

 State Space Representation

𝑑$𝑥
𝑀 $ + 𝐾𝒙 𝑡 = 𝐹(𝑡)
𝑑𝑡
Repacking 1st order System
𝑋 𝑠 1
= 𝑥̇ = 𝑓(𝑥, 𝑢)
𝐹 𝑠 𝑀𝑠 $ + 𝐾
Defines how system changes with
its state and input
7/10/22 10:52 PM SK, Pulchowk Campus 18
 State Space Representation
State Equation
𝑋̇ = 𝐴𝑋 + 𝐵𝑈
Output Equation
𝑌 = 𝐶𝑋 + 𝐷𝑈
where,
𝑋̇ = derivative of state variable vector
𝑋 = state vector
𝑈 = input vector
𝑌 = output vector
𝐴, 𝐵, 𝐶, 𝐷 = transition matrix
Aà State Matrix of the System[Characteristic Property of the system]
Bà Input Matrix [Parameters related to scaling Input]
Cà Output Matrix [Relating State to Output]
Dà Feed forward Matrix
7/10/22 10:52 PM SK, Pulchowk Campus 19
 Elements in Electrical System

7/10/22 10:52 PM SK, Pulchowk Campus 20

 Elements in Electrical System

7/10/22 10:53 PM SK, Pulchowk Campus 21

 Steps for State Space Representation
1. Identify the number of energy storage elements i.e. Integral that will be present
in the system. [It gives the order of the system]
2. Identify the variables that will be used for State Space Representation.
q System Response is dependent on
i. Input [u]
ii. Integrator Output Value [x]
q Find all us and Xs
3. Starting with dynamic elements, Obtain differential equations
Ø No of Differential Equation = Order of Equation
Ø No of Differential Equation =No of Energy Storage Elements/Memory
Elements
4. Segregating into the state equation form
𝑋̇ = 𝐴𝑋 + 𝐵𝑈
5. Segregating into the output equation form
𝑌 = 𝐶𝑋 + 𝐷𝑈
7/10/22 10:53 PM SK, Pulchowk Campus 22
State Representation of Electrical System: Example
1. No. of Storage Element =1, So, Order of the system is
one.
2. We have capacitor So,
1
𝑉& = J 𝑖 𝑑𝑡
𝐶
So, State Variable is 𝑉& and Input Variable is 𝑉'%
3. No of Differential equation=1
1
𝑉& = J 𝑖 𝑑𝑡
𝐶
1 𝑉'% − 𝑉!()
𝑉& = J 𝑑𝑡 𝑉!() = 𝑉&
𝐶 𝑅
𝑑𝑉& 1 𝑉'% − 𝑉!() 1
= = {𝑉'% − 𝑉& }
𝑑𝑡 𝐶 𝑅 𝐶𝑅
7/10/22 10:53 PM SK, Pulchowk Campus 23
State Representation of Electrical System : Example
𝑑𝑉& 1
= {𝑉'% − 𝑉& }
𝑑𝑡 𝐶𝑅
4. Segregating in State Equation Form,
𝑑𝑉& 1 1
=− 𝑉& + 𝑉'%
𝑑𝑡 𝐶𝑅 𝐶𝑅
1 1
𝑉&̇ = − 𝑉& + 𝑉'%
𝐶𝑅 𝐶𝑅
5. Segregating in Output Equation Form,
𝑉!() = 1 𝑉& + [0]. 𝑉'%

State Space representation has been achieved.

7/10/22 10:53 PM SK, Pulchowk Campus 24

State Representation of Electrical System : Example
1. No. of Storage Element =2, So, Order of the system is
two.
2. Two storage elements, So, State Variable is 𝑉& and 𝑖*
and Input Variable is 𝑉' .
𝑉&
𝑥 =
𝑖*
𝑢 = [𝑉' ]
System Parameters are: 𝑅#, 𝐿, 𝐶 , 𝑅$
3. No of Differential equation=2. Starting with dynamic elements.
For Inductor,
Apply KVL in first loop.
𝑉' − 𝑉+# − 𝑉* − 𝑉, = 0 𝑑𝑖* 𝑉' − 𝑖* 𝑅# − 𝑉,
=
𝑉* = 𝑉' − 𝑉+# − 𝑉, 𝑑𝑡 𝐿
𝑑𝑖* 𝑑𝑖* −𝑖* 𝑅# − 𝑉, V-
𝐿 = 𝑉' − 𝑖* 𝑅# − 𝑉, = +
𝑑𝑡 𝑑𝑡 𝐿 L
7/10/22 10:53 PM SK, Pulchowk Campus 25
State Representation of Electrical System : Example
For Capacitor,
𝑑𝑉&
𝐶 = 𝑖* − 𝑖"
𝑑𝑡
𝑑𝑉& 𝑉&
𝐶 = 𝑖* −
𝑑𝑡 𝑅$
𝑉! = 𝑉& 𝑑𝑉& 𝑖* 𝑉&
= −
𝑑𝑡 𝐶 𝑅$𝐶

4. Segregating in to State equation 4. Segregating in to State equation

form form
𝑑𝑖* −𝑖* 𝑅# − 𝑉, V- 𝑑𝑉& 𝑖* 𝑉&
= + = −
𝑑𝑡 𝐿 L 𝑑𝑡 𝐶 𝑅$𝐶
𝑅 # 1 𝑖 * 1 1 1 𝑖*
𝑖*̇ = − − + 𝑉' ̇
𝑉& = − + 0 𝑉'
𝐿 𝐿 &𝑉 𝐿 𝐶 𝑅$𝐶 & 𝑉

7/10/22 10:53 PM SK, Pulchowk Campus 26

 State Representation of Electrical System : Example
𝑅# 1 𝑖* 1
̇
𝑖* = − − + 𝑉'
𝐿 𝐿 𝑉& 𝐿
1 1 𝑖*
𝑉&̇ = − + 0 𝑉'
𝐶 𝑅$𝐶 𝑉&
In Combined Form,
𝑅# 1
− − 1
𝑖*̇ 𝐿 𝐿 𝑖*
5. Segregating in to Output equation = 1 1 + 𝐿 [𝑉' ]
𝑉&̇ 𝑉&
form − 0
𝐶 𝑅$𝐶
𝑉! = 𝑉&
𝑖*
𝑉" = 0 1 + 0 𝑉'
𝑉&

7/10/22 10:53 PM SK, Pulchowk Campus 27

State Representation of Electrical System : Example[6]
Note: While writing state equation, use KCL at
node which contain capacitor & KVL in loop
containing inductor

7/10/22 10:53 PM SK, Pulchowk Campus 28

State Representation of Mechanical System : Example
1. No of state variables is two times the No of Mass or
displacement.
2. Two storage elements, So, State Variable is 𝑉& and 𝑖*
and Input Variable is 𝑉' .
𝑥 𝑥#
𝑥 = = 𝑥
𝑣 = 𝑥̇ $
𝑢 = [𝐹]
System Parameters are: 𝑘, 𝐵 𝑎𝑛𝑑 𝑚
3. No of Differential equation=2.
𝑥#̇ = 𝑥$
From Freebody Balance approach,
𝑑$𝑥 𝑑𝑥
𝑚 $ +𝐵 + 𝑘𝑥 = 𝐹 𝑚𝑥$̇ + 𝐵𝑥$ + 𝑘𝑥# = 𝐹
𝑑𝑡 𝑑𝑡
𝑚𝑥̈ + 𝐵𝑥̇ + 𝑘𝑥 = 𝐹 𝐵 𝑘 𝐹
𝑥$̇ = − 𝑥$ − 𝑥# +
𝑚𝑥$̇ + 𝐵𝑥$ + 𝑘𝑥# = 𝐹 𝑚 𝑚 𝑚
7/10/22 10:53 PM SK, Pulchowk Campus 29
State Representation of Mechanical System : Example
𝑥#̇ = 𝑥$
𝐵 𝑘 𝐹
𝑥$̇ = − 𝑥$ − 𝑥# +
𝑚 𝑚 𝑚
Rewriting in matrix form,
0 1 𝑥# 0
𝑥#̇
= 𝑘 𝐵
𝑥 + 1 [𝐹]
𝑥$̇ − − $
𝑚 𝑚 𝑚
Lets Consider, we are looking for position as Output,
𝑥#
𝑦= 1 0 𝑥
$

7/10/22 10:53 PM SK, Pulchowk Campus 30

State Representation of Mechanical System : Example

7/10/22 10:53 PM SK, Pulchowk Campus 31

 Find the steady state representation
𝑑.𝑦 𝑑$𝑦 𝑑𝑦
. + 6 $ + 11 + 10𝑦 = 8𝑢(𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡

𝑥#̇ 0 1 0 𝑥# 0
𝑥$̇ = 0 0 1 𝑥$ + 0 𝑢(𝑡)
𝑥.̇ −10 −11 −6 𝑥. 8
7/10/22 10:53 PM SK, Pulchowk Campus 32
 Syllabus
8.1. Definition of state -space
8.2. State space representation of electrical and mechanical system
8.3. Conversion from state space to transfer function
8.4. Conversion from transfer function to state space
8.5. State-transition matrix

7/10/22 10:53 PM SK, Pulchowk Campus 33

Conversion from State Space to Transfer Function
We have,
𝑋̇ = 𝐴𝑋 + 𝐵𝑈
𝑌 = 𝐶𝑋 + 𝐷𝑈
Laplace Transform on both sides,
𝑠𝑋 𝑠 = 𝐴 𝑋 𝑠 + 𝐵 𝑈(𝑠)
𝑠𝐼 − 𝐴 𝑋 𝑠 = 𝐵 𝑈(𝑠)
𝑋 𝑠 = 𝑠𝐼 − 𝐴 !" 𝐵𝑈(𝑠)

𝑌 𝑠 = 𝐶 𝑋 𝑠 + 𝐷 𝑈(𝑠)
= 𝐶 𝑠𝐼 − 𝐴 !" 𝐵𝑈(𝑠) + 𝐷 𝑈(𝑠)
𝑌 𝑠
= 𝐶 𝑠𝐼 − 𝐴 !" 𝐵 + 𝐷
𝑈(𝑠)
𝒀 𝒔
∴ 𝑻. 𝑭 = 𝑮 𝒔 = = 𝑪 𝒔𝑰 − 𝑨 !𝟏 𝑩 + 𝑫
𝑼(𝒔)
where,
!"
𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝑜𝑓 𝑠𝐼 − 𝐴 𝑇𝑟𝑎𝑛𝑠𝑝𝑜𝑠𝑒 𝑜𝑓 𝑚𝑎𝑡𝑟𝑖𝑥 𝑜𝑓 𝑐𝑜_𝑓𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 [𝑠𝐼 − 𝐴]
𝑠𝐼 − 𝐴 = =
|𝑠𝐼 − 𝐴| |𝑠𝐼 − 𝐴|
7/10/22 10:53 PM SK, Pulchowk Campus 34
 Example
Obtain the transfer function from the given
steady state model 𝑠+5 1
𝑥#̇ 0 1 𝑥# 0 /𝟏 −6 𝑠 0
= + 𝑢 𝑡 𝑪 𝑩 𝒔𝑰 − 𝑨 = 8 1 $
𝑥$̇ −6 −5 $ 𝑥 1 𝑠 + 5𝑠 + 6 1
𝑥# 𝐬+𝟖
𝑦= 8 1 𝑥 𝐆 𝐬 = 𝟐
$ 𝐬 + 𝟓𝐬 + 𝟔
𝒀 𝒔
𝑮 𝒔 = = 𝑪 𝒔𝑰 − 𝑨 /𝟏 𝑩 + 𝑫
𝑼(𝒔)
1 0 0 1
𝑠𝐼 − 𝐴 = 𝑠 −
0 1 −6 −5
𝑠 −1 Result Can be
= verified by
6 𝑠+5 checking poles
/#
𝑎𝑑𝑗𝑜𝑖𝑛𝑡 𝑜𝑓 [𝑠𝐼 − 𝐴] with eigen
𝑠𝐼 − 𝐴 =
|𝑠𝐼 − 𝐴| values.
𝑠+5 1
𝑠𝐼 − 𝐴 /# = $ −6 𝑠
7/10/22 10:53 PM 𝑠 + 5𝑠 + 6 SK, Pulchowk Campus 35
 Lets Check Stability from State Space Representation
𝒀 𝒔
∴ 𝑻. 𝑭 = 𝑮 𝒔 = = 𝑪 𝒔𝑰 − 𝑨 !𝟏 𝑩 + 𝑫
𝑼(𝒔)
𝒀 𝒔 𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝑜𝑓 𝑠𝐼 − 𝐴
∴ 𝑻. 𝑭 = 𝑮 𝒔 = =𝑪 𝑩 +𝑫
𝑼(𝒔) |𝑠𝐼 − 𝐴|
So, the Characteristic equation is, 𝑠𝐼 − 𝐴 = 0
On solving that, we get poles and based on its location stability can be predicted.

7/10/22 10:53 PM SK, Pulchowk Campus 36

 Example
Check the stability of the system
𝑥#̇ 0 1 𝑥# 0
= 𝑥 + 𝑢 𝑡
𝑥$̇ −6 −5 $ 1
𝑥#
𝑦= 8 1 𝑥
$
𝒀 𝒔
𝑮 𝒔 = = 𝑪 𝒔𝑰 − 𝑨 /𝟏 𝑩 + 𝑫
𝑼(𝒔)
1 0 0 1 𝑠 −1
𝑠𝐼 − 𝐴 = 𝑠 − =
0 1 −6 −5 6 𝑠+5
$
|𝑠𝐼 − 𝐴| = 𝑠 + 5𝑠 + 6
So, Characteristic Equation is, 𝑠 $ + 5𝑠 + 6 = 0
Poles are: 𝑠 = −2 𝑎𝑛𝑑 𝑠 = −3
Both are on LHS Side, so system is stable.

7/10/22 10:53 PM SK, Pulchowk Campus 37

 Syllabus
v8.1.
Definition of state -space
v8.2. State space representation of electrical and mechanical system
v8.3. Conversion from state space to transfer function
v8.4. Conversion from transfer function to state space
v 8.4.1) T.F containing poles only
v 8.4.2) T.F containing both poles and zeros
v8.5. State-transition matrix

7/10/22 10:53 PM SK, Pulchowk Campus 38

Steps for Conversion from TF to SS: Only Poles
𝑌 𝑠
𝑇. 𝐹 =
𝑈 𝑠
1. Cross multiply
2. Take inverse Laplace Transform
3. Replace with state variables
For example
for system order = 3 define state variables i.e 𝑥#, 𝑥$ & 𝑥. as follows:
Let,
𝑦 = 𝑥#
𝑦̇ = 𝑥#̇ = 𝑥$
𝑦̈ = 𝑥̈ # = 𝑥$̇ = 𝑥.
𝑦⃛ = 𝑥⃛# = 𝑥$̈ = 𝑥.̇
4. Get state equation & output equation
7/10/22 10:53 PM SK, Pulchowk Campus 39
 Example
𝒀 𝒔 𝟐
𝐐)𝐅𝐢𝐧𝐝 𝐬𝐭𝐚𝐭𝐞 𝐞𝐪𝐮𝐚𝐭𝐢𝐨𝐧 𝒇𝐨𝐫 𝐓. 𝐅 = = 𝟑 𝟐 .
𝑼 𝒔 𝒔 + 𝟔𝒔 + 𝟏𝟏𝒔 + 𝟔

We have,
𝑌 𝑠 2
= .
𝑈 𝑠 𝑠 + 6𝑠 $ + 11𝑠 + 6
𝑜𝑟, 𝑠 . 𝑌 𝑠 + 6 𝑠 $ 𝑌 𝑠 + 11 𝑠 𝑌 𝑠 + 6 𝑌 𝑠 = 2 𝑈(𝑠)
Taking inverse L.T
𝑦⃛ 𝑡 + 6𝑦̈ 𝑡 + 11𝑦̇ 𝑡 + 6𝑦 𝑡 = 2𝑢(𝑡)
𝑦⃛ 𝑡 = 2𝑢 𝑡 − 6𝑦 𝑡 − 11𝑦̇ 𝑡 − 6𝑦̈ 𝑡 … … … … … (𝑖)

7/10/22 10:53 PM SK, Pulchowk Campus 40

 Example
Let, 𝑦 = 𝑥# Therefore the state equations are :
𝑦̇ = 𝑥#̇ = 𝑥$ [ Note: 𝑋̇ = 𝐴𝑋 + 𝐵𝑈]
𝑦̈ = 𝑥̈ # = 𝑥$̇ = 𝑥. 𝑥̇ # = 𝑥$ = 0𝑥# + 1𝑥$ + 0𝑥. + 0 𝑢(𝑡)
Since the order of equation is three, so 𝑥̇ $ = 𝑥. = 0𝑥# + 0𝑥$ + 1𝑥. + 0 𝑢(𝑡)
there are 3 state variables 𝑥# , 𝑥$ , 𝑥.. 𝑥.̇ = −6𝑥# − 11𝑥$ − 6𝑥. + 2𝑢 𝑡
𝑦⃛ = 𝑥⃛# = 𝑥$̈ = 𝑥.̇ &
From (i) we have, The output equation is: [Note: 𝑌 = 𝐶𝑋 + 𝐷𝑈 ]
𝑥.̇ = 2𝑢 𝑡 − 6𝑥# − 11𝑥$ − 6𝑥. 𝑦 = 𝑥# = 𝑥# + 0𝑥$ + 0𝑥. + 0 𝑢(𝑡)

7/10/22 10:53 PM SK, Pulchowk Campus 41

 Example
The state space model becomes:
𝑋̇ = 𝐴𝑋 + 𝐵𝑈
𝑥̇ # 0 1 0 𝑥# 0
𝑥̇ $ = 0 0 1 𝑥$ + 0 𝑢(𝑡)
𝑥.̇ −6 −11 −6 𝑥. 2
&
𝑌 = 𝐶𝑋 + 𝐷𝑈
𝑥#
= 1 0 0 𝑥$ + 0 𝑢(𝑡)
𝑥.

7/10/22 10:53 PM SK, Pulchowk Campus 42

 Syllabus
v8.1.
Definition of state -space
v8.2. State space representation of electrical and mechanical system
v8.3. Conversion from state space to transfer function
v8.4. Conversion from transfer function to state space
v 8.4.1) T.F containing poles only
v 8.4.2) T.F containing both poles and zeros
v8.5. State-transition matrix

7/10/22 10:53 PM SK, Pulchowk Campus 43

 Steps for Conversion from TF to SS
𝒀 𝒔 𝑿𝟏 𝒔 𝒀 𝒔
𝑮 𝒔 = = ∗
𝑼(𝒔) 𝑼(𝒔) 𝑿𝟏 𝒔
𝑿𝟏 𝒔 𝒀 𝒔
= 𝒑𝒐𝒍𝒆 𝒕𝒆𝒓𝒎
𝑼(𝒔) = 𝒛𝒆𝒓𝒐 𝒕𝒆𝒓𝒎
𝑿𝟏 𝒔
1. Cross multiply
2. Take inverse Laplace Transform 1. Cross multiply
3. Replace with state variables 2. Take inverse Laplace Transform
For example 3. Replace with state variables
for system order = 3 define state variables
i.e 𝑧" , 𝑧$ & 𝑧% as follows:
4. Get output equation
Let,
𝑥" 𝑡 = 𝑧"
𝑥"̇ 𝑡 = 𝑧"̇ = 𝑧$
𝑥"̈ 𝑡 = 𝑧"̈ = 𝑧$̇ = 𝑧%
𝑥⃛" 𝑡 = 𝑧⃛" = 𝑧"̈ = 𝑧%̇
4. Get state equation
7/10/22 10:53 PM SK, Pulchowk Campus 44
 Example
𝟐𝟎 𝟏𝟎𝒔 + 𝟏 From (𝑖)
𝐀 𝐬𝐲𝐬𝐭𝐞𝐦 𝐡𝐚𝐬 𝐓. 𝐅 𝑮 𝒔 = 𝟑 .
𝒔 + 𝟑𝒔𝟐 + 𝟐𝒔 + 𝟏
𝐅𝐢𝐧𝐝 𝐭𝐡𝐞 𝐬𝐭𝐚𝐭𝐞 & 𝐨𝐮𝐭𝐩𝐮𝐭 𝐞𝐪𝐮𝐚𝐭𝐢𝐨𝐧 𝐢𝐧 𝐦𝐚𝐭𝐫𝐢𝐱 𝐟𝐨𝐫𝐦.
𝑠 .𝑋# 𝑠 + 3𝑠 $𝑋# 𝑠 + 2𝑠𝑋# 𝑠 + 𝑋# 𝑠
Solution,
= 20 𝑈(𝑠)
𝑌 𝑠 20 Inverse L.T
𝐺 𝑠 = = ∗ (10𝑠 + 1)
𝑈(𝑠) 𝑠 % + 3𝑠 $ + 2𝑠 + 1 𝑥⃛# 𝑡 + 3𝑥#̈ 𝑡 + 2𝑥#̇ 𝑡 + 𝑥# 𝑡 = 20 𝑢(𝑡)
𝑌 𝑠 𝑋" 𝑠 𝑌 𝑠 𝑥⃛# 𝑡 = 20 𝑢 𝑡 − 𝑥# 𝑡 − 2𝑥#̇ 𝑡 − 3𝑥#̈ 𝑡
= ∗
𝑈(𝑠) 𝑈(𝑠) 𝑋" 𝑠
So, Let, 𝑥# 𝑡 = 𝑧#
𝑋" 𝑠 20 𝑥#̇ 𝑡 = 𝑧#̇ = 𝑧$
= … … … … … . (𝑖)
𝑈(𝑠) 𝑠 % + 3𝑠 $ + 2𝑠 + 1 𝑥#̈ 𝑡 = 𝑧#̈ = 𝑧$̇ = 𝑧.
𝑌 𝑠 𝑥⃛# 𝑡 = 20 𝑢 𝑡 − 𝑧# − 2𝑧$ − 3𝑧.
= 10𝑠 + 1 … … … … . (𝑖𝑖)
𝑋" 𝑠 𝑥⃛# 𝑡 = 𝑧⃛# = 𝑧#̈ = 𝑧.̇

7/10/22 10:53 PM SK, Pulchowk Campus 45

 Example
The state equations in form 𝑍̇ = 𝐴𝑍 + 𝐵𝑈
∴ The state matrix is as follows:
𝑧#̇ = 𝑧$ = 0𝑧# + 1𝑧$ + 𝑧. + 0𝑢(𝑡)
𝑧$̇ = 𝑧. = 0𝑧# + 0𝑧$ + 1𝑧. + 0𝑢(𝑡) 𝑧#̇ 0 1 0 𝑧# 0
𝑧.̇ = −𝑧# − 2𝑧$ − 3𝑧. + 20 𝑢 𝑡 𝑧$̇ = 0 0 1 𝑧$ + 0 𝑢(𝑡)
&
𝑧.̇ −1 −2 −3 𝑧. 20
which is in the form 𝑍̇ = 𝐴𝑍 + 𝐵𝑈.
The output equation in form 𝑌 = 𝐶𝑍 + 𝐷𝑈
&
From (ii)
𝑌 𝑠 = 10𝑠𝑋# 𝑠 + 𝑋#(𝑠) 𝑧#
𝑌 = 1 10 0 𝑧$
Taking inverse L.T 𝑧.
𝑦 𝑡 = 10𝑥#̇ 𝑡 + 𝑥# 𝑡 = 10𝑧$ + 𝑧# which is in the form 𝑌 = 𝐶𝑍 + 𝐷𝑈.
𝑦 𝑡 = 1𝑧# + 10𝑧$ + 0𝑧. + 0 𝑢(𝑡)

7/10/22 10:53 PM SK, Pulchowk Campus 46

 Controllability and Observability

𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑙𝑒𝑟 𝑄𝑢𝑎𝑑𝑜𝑝𝑡𝑒𝑟
-

𝑆𝑒𝑛𝑠𝑜𝑟

𝑿̇ = 𝐴x + Bu
𝑌 = 𝐶𝒙

7/10/22 10:53 PM SK, Pulchowk Campus 47

 Controllability : A Small Math
𝑋̇ = 𝐴x + Bu 𝑋̇ = 𝐴𝑋 + 𝐵𝑢
𝑌 = Cx
Lets assume, 𝑌 = 𝑥
The Sensor is feeding back the output,
then
So, −𝐾𝑋
𝑋̇ = 𝐴x − BKx = [A − BK]x
For Controllability, I can achieve any Control System
Optimal for Linear
value by changing the value of K but
that comes in picture when I can
control the u such that to change the
system.

7/10/22 10:53 PM SK, Pulchowk Campus 48

 Is the System Controllable or Not?
𝑥#̇ 2 0 𝑥# 0 𝑥#̇ 2 0 𝑥# 1 0 𝑢#
= + 𝑢 = 𝑥 +
𝑥$̇ 𝑥
0 1 $ 1 𝑥$̇ 0 1 $ 0 1 𝑢$

𝑥#̇ 2 4 𝑥# 0 𝑋̇ = 𝐴x + Bu
= 𝑥 + 𝑢 The Solution will be,
𝑥$̇ 0 1 $ 1
)
𝑥 𝑡 = 𝑒 I) 𝑥 0 + J 𝑒 I )/J 𝐵𝑢 𝜏 𝑑𝜏
"

7/10/22 10:53 PM SK, Pulchowk Campus 49

 Controllability Matrix
C = 𝐵 ∶ 𝐴𝐵 ∶ 𝐴$𝐵 … … . 𝐴%/#𝐵
If rank(C)=n,
the system is controllable.

v A control system is said to be controllable if the initial states of the control

system are transferred to some other desired states by a controlled input in
finite duration of time.
v We can check the controllability of a control system by using Kalman’s test.

7/10/22 10:53 PM SK, Pulchowk Campus 50

 Is the System Controllable or Not?
𝑥#̇ 2 0 𝑥# 0 𝑥#̇ 2 4 𝑥# 0
= 𝑥 + 𝑢 = 𝑥 + 𝑢
𝑥$̇ 0 1 $ 1 𝑥$̇ 0 1 $ 1

C = 𝐵 ∶ 𝐴𝐵 ∶ 𝐴$𝐵 … … . 𝐴%/#𝐵 C = 𝐵 ∶ 𝐴𝐵 ∶ 𝐴$𝐵 … … . 𝐴%/#𝐵

0 0 0 4
C= C=
1 1 1 1
𝑅𝑎𝑛𝑘 = 1 ≠ 2 𝑅𝑎𝑛𝑘 = 2 = 2
So, System is Uncontrollable System So, System is Controllable System

7/10/22 10:53 PM SK, Pulchowk Campus 51

 How much controllability? Extra
It can be found by looking at eigen values/vectors of Gramian of Controllability
)
IJ K I!J
𝑊) = J 𝑒 𝐵𝐵 𝑒 𝑑𝜏 ∈ 𝔑%×%
"
𝑊) ≈ 𝒞×𝒞 K
≫ 𝑈, 𝑆, 𝑉 = 𝑠𝑣𝑑(𝒞, ′𝑒𝑐𝑜𝑛′)

7/10/22 10:53 PM SK, Pulchowk Campus 52

 Stabilizability
For partially controllable system, if the uncontrollable modes are stable and the
other modes are controllable, then the system is said to be stabilizable.

7/10/22 10:53 PM SK, Pulchowk Campus 53

 Observability Matrix
𝐶
𝐶𝐴
𝒪= 𝐶𝐴 $
…
𝐶𝐴%/#
If rank(𝒪)=n,
the system is Observable.

v A control system is said to be observable if it is able to determine the initial

states of the control system by observing the outputs in finite duration of
time.
v We can check the Observable of a control system by using Kalman’s test.

7/10/22 10:53 PM SK, Pulchowk Campus 54

 Lets Check Observability
0 1 0 𝐶
𝐴= 0 0 1 𝐶𝐴
−1 −2 −3 𝒪= 𝐶𝐴 $
0 …
𝐵= 0 𝐶𝐴%/#
20
𝐶 = 1 10 0 1 10 0
𝒪= 0 1 10
−10 −20 −29
𝑅𝑎𝑛𝑘 = 3 = 𝑛
So, it is Observable.

7/10/22 10:53 PM SK, Pulchowk Campus 55

 How much Observability?
It can be found by looking at eigen values/vectors of Gramian of Observability
)
I!J K
𝑊" = J 𝑒 𝐶 𝐶𝑒 IJ 𝑑𝜏 ∈ 𝔑%×%
"
≫ 𝑈, 𝑆, 𝑉 = 𝑠𝑣𝑑(𝒪)

7/10/22 10:53 PM SK, Pulchowk Campus 56

 Syllabus
8.1.Definition of state -space
8.2. State space representation of electrical and mechanical system
8.3. Conversion from state space to transfer function
8.4. Conversion from transfer function to state space
8.4.1) T.F containing poles only
8.4.2) T.F containing both poles and zeros
8.5. State-transition matrix

7/10/22 10:53 PM SK, Pulchowk Campus 57

 Homogeneous Equation and Non-Homogeneous Equation
𝑋̇ = 𝐴x + Bu
v If A is a constant matrix and input control forces are zero, then such equation is
called homogenous equation.
𝑋̇ = 𝐴x
v The new state is due to their initial conditions.
v If both are present,.i.e.
𝑋̇ = 𝐴x + Bu
v Such equation is called non-homogeneous equation.

7/10/22 10:53 PM SK, Pulchowk Campus 58

Lets Find the solution for Homogeneous Equation

7/10/22 10:53 PM SK, Pulchowk Campus 59

Lets Find the solution for Homogeneous Equation

7/10/22 10:53 PM SK, Pulchowk Campus 60

Lets Find the solution for Homogeneous Equation

7/10/22 10:53 PM SK, Pulchowk Campus 61

Lets Find the solution for Homogeneous Equation

7/10/22 10:53 PM SK, Pulchowk Campus 62

 State Transition Matrix
v It can be observed that without input, initial state X(0) drives the state X(t) at
any time t.
v Thus there is transition of the initial state X(0) from initial time t = 0 to any
time t through the matrix eAt.
v As it is an exponential term, the matrix eAt is called matrix exponential.
v It is also responsible for the transition of the state X(t) at any time t from initial
time hence also called state transition matrix.
v It is denoted as 𝜙(𝑡).
𝜙 𝑡 = e I) = 𝑆𝑡𝑎𝑡𝑒 𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖𝑜𝑛 𝑀𝑎𝑡𝑟𝑖𝑥
v If we shift from t0
𝜙 𝑡 = e I()/)" )

7/10/22 10:53 PM SK, Pulchowk Campus 63

 State Transition Matrix: By Laplace
We know, Let,
𝑋̇ 𝑡 = 𝐴𝑋 𝑡 + 𝐵𝑈(𝑡) 𝜙 𝑠 = 𝑠𝐼 − 𝐴 /#
= Resolvent matrix
If 𝑈 𝑡 = 0,
𝑋̇ 𝑡 = 𝐴𝑋 𝑡
Taking inverse L.T,
Taking L.T. 𝜙 𝑡 = 𝐿/# 𝜙 𝑠
𝑠𝑋 𝑠 − 𝑋 0 = 𝐴𝑋(𝑠) = 𝐿/#{ 𝑠𝐼 − 𝐴 /# }
𝑠𝐼 − 𝐴 𝑋 𝑠 = 𝑋(0)
𝑋 𝑠 = 𝑠𝐼 − 𝐴 /#𝑋(0)

𝝓 𝒕 = 𝒆𝑨𝒕 = 𝑺. 𝑻. 𝑴 = 𝑳/𝟏 { 𝒔𝑰 − 𝑨 /𝟏 }

7/10/22 10:53 PM SK, Pulchowk Campus 64

 Computation of State Transition Matrix
v While obtaining the solution of state equation, the computation of state
transition matrix eAt plays an important role.
v There are various methods of obtaining state transition matrix from the state
model.
v These methods are,
Ø Laplace transform method
Ø Power series method
Ø Cayley Hamilton method
Ø Similarity transformation method

7/10/22 10:53 PM SK, Pulchowk Campus 65

 Computation of State Transition Matrix
v While obtaining the solution of state equation, the computation of state
transition matrix eAt plays an important role.
v There are various methods of obtaining state transition matrix from the state
model.
v These methods are,
Ø Laplace transform method
Ø Power series method
Ø Cayley Hamilton method
Ø Similarity transformation method

7/10/22 10:53 PM SK, Pulchowk Campus 66

 Computation of State Transition Matrix
1) Write state space model
2) Find 𝑠𝐼 − 𝐴 /#
/#
𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝑜𝑓 𝑠𝐼 − 𝐴
𝑠𝐼 − 𝐴 =
𝑠𝐼 − 𝐴
where,
𝐼 = 𝐼𝑑𝑒𝑛𝑡𝑖𝑡𝑦 𝑚𝑎𝑡𝑟𝑖𝑥
𝑠𝐼 − 𝐴 = 𝑑𝑒𝑡𝑒𝑟𝑚𝑖𝑛𝑎𝑛𝑡 𝑜𝑓 𝑠𝐼 − 𝐴
3) State transition matrix (STM) = 𝐿/# 𝑠𝐼 − 𝐴 /# = 𝑒 I)
4) The Solution of Homogeneous equation is, 𝑋(𝑡) = 𝑒 I) 𝑋(0)

7/10/22 10:53 PM SK, Pulchowk Campus 67

 Computation of State Transition Matrix
Evaluate the state transition matrix of system represented by:
𝒙̇ 𝟏 = −𝒙𝟏 + 𝒙𝟐
𝒙̇ 𝟐 = −𝟐𝒙𝟐
𝟏
Find Solution of Homogeneous equation if 𝑿 𝟎 =
𝟎
Solution,
The state space model is:
𝑥̇ # −1 1 𝑥#
=
𝑥̇ $ 0 −2 𝑥$
1 0 −1 1
𝑠𝐼 − 𝐴 = 𝑠 −
0 1 0 −2
𝑠 + 1 −1
=
0 𝑠+2
𝑠𝐼 − 𝐴 = 𝑠 + 1 𝑠 + 2

7/10/22 10:53 PM SK, Pulchowk Campus 68

 Computation of State Transition Matrix
/#
𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝑜𝑓 𝑠𝐼 − 𝐴 State transition matrix (STM)
𝑠𝐼 − 𝐴 = = 𝐿/# 𝑠𝐼 − 𝐴 /# = 𝑒 I)
𝑠𝐼 − 𝐴
1 𝑠+2 1 /#
1 /#
1
= 𝐿 𝐿
𝑠+1 𝑠+2 0 𝑠+1 𝑠+1 𝑠+1 𝑠+2
1 1 =
/#
1
0 𝐿
𝑠+1 𝑠+1 𝑠+2 𝑠+2
= /) 𝑒 /) − 𝑒 /$)
1 𝑒
0 =
𝑠+2 0 𝑒 /$)

𝑋 𝑡 = 𝑒 I) 𝑋 0 = 𝑍𝑒𝑟𝑜 𝐼𝑛𝑝𝑢𝑡 𝑅𝑒𝑠𝑝𝑜𝑛𝑠𝑒

𝑒 /) 𝑒 /) − 𝑒 /$) 1 𝑒 /)
𝑋 𝑡 = /$) =
0 𝑒 0 0
7/10/22 10:53 PM SK, Pulchowk Campus 69
 Thank You
???
References
1. S. Bhattarai, “ Control System”, TU, IOE.
2. Nagoor Kani, “Control System”
3. Nise, Norman S. "Control system engineering, John Wiley & Sons." Inc, New York (2011).
4. Dorf, Richard C., and Robert H. Bishop. Modern control systems. Pearson, 2011.
5. Bakshi, Uday A., and Mayuresh V. Bakshi. Modern control theory. Technical Publications, 2020.
6. Ogata, Katsuhiko. Modern control engineering. Prentice hall, 2010.
7. Gopal, Madan. Control systems: principles and design. Tata McGraw-Hill Education, 2002.
8. Dr. RamKrishna Pasumarthy, “Control System Engineering”, NPTEL
Assignment : 1: Represent the system in state space form

7/10/22 10:53 PM SK, Pulchowk Campus 71

Assignment 2: Find State space model for speed control

7/10/22 10:53 PM SK, Pulchowk Campus 72

 Example
Q) Find T.F of the following system. The co-factor matrix of [𝑠𝐼 − 𝐴] be:

𝑎## 𝑎#! 𝑎#"

𝟎 𝟏 𝟎 𝟎 𝑎!# 𝑎!! 𝑎!"
̇𝑿 𝒕 = −𝟏 −𝟏 𝟎 𝑿 + 𝟏 𝑼 & 𝒀 = 𝟎 𝟎 𝟏 𝑿 Note: 𝑎$% = omit 𝑖 &' row & 𝑗&' column from [𝑠𝐼 − 𝐴]
𝑎"# 𝑎"! 𝑎""
𝟏 𝟎 𝟎 𝟎
So,
Solution,

We know, 𝑠+1 0
𝑎## = =𝑠 𝑠+1
0 𝑠
𝑋̇ 𝑡 = 𝐴 𝑋 𝑡 + 𝐵 𝑈 𝑡 à State equation
1 0
𝑎#! = − = −𝑠
−1 𝑠
𝑌 𝑡 = 𝐶 𝑋 𝑡 + 𝐷 𝑈(𝑡) à Output equation

Comparing, 1 𝑠+1
𝑎#" = = (𝑠 + 1)
−1 0
0 1 0 0
𝐴 = −1 −1 0 ,𝐵 = 1 ,𝐶 = 0 0 1 , 𝐷 = [0] −1 0
𝑎!# = = −𝑠
1 0 0 0 1 𝑠

Now, 𝑠 0
𝑎!! = = 𝑠!
−1 𝑠
1 0 0 0 1 0 𝑠 −1 0
𝑠𝐼 − 𝐴 = 𝑠 0 1 0 − −1 −1 0 = 1 𝑠+1 0 𝑠 −1
𝑎!" = =1
0 0 1 1 0 0 −1 0 𝑠 −1 0
𝑠+1 0 1 0 Similarly, 𝑎"# = 0 , 𝑎"! = 0 , & 𝑎"" = 𝑠 ! + 𝑠 + 1
𝑠𝐼 − 𝐴 = 𝑠 − −1 +0
0 𝑠 −1 𝑠

= s! 𝑠 + 1 + 𝑠 𝑠(𝑠 + 1) −𝑠 𝑠+1
∴ Matrix of co − factors = 𝑠 𝑠! 1
= 𝑠" + 𝑠! + 𝑠 0 0 𝑠! + 𝑠 + 1

7/10/22 10:53 PM SK, Pulchowk Campus 73

 Example
∴ Adjoint of 𝑠𝐼 − 𝐴 = Transponse of matrix of co − factors
𝑠(𝑠 + 1) 𝑠 0
= −𝑠 𝑠+ 0
𝑠+1 1 𝑠+ + 𝑠 + 1
Adjoint of 𝑠𝐼 − 𝐴 1 𝑠 𝑠+1 𝑠 0
,-
∴ 𝑠𝐼 − 𝐴 = = . −𝑠 𝑠+ 0
𝑠𝐼 − 𝐴 𝑠 + 𝑠+ + 𝑠
𝑠+1 1 𝑠+ + 𝑠 + 1
,-
∴ 𝑇. 𝐹 = 𝐶 𝑠𝐼 − 𝐴 𝐵+𝐷

1 𝑠 𝑠+1 𝑠 0 0
= 0 0 1 ∗ . −𝑠 𝑠+ 0 ∗ 1 +0
𝑠 + 𝑠+ + 𝑠
𝑠+1 1 𝑠+ + 𝑠 + 1 0
1 𝑠
+
= 0 0 1 ∗ . ∗ 𝑠
𝑠 + 𝑠+ + 𝑠 1
1
=
𝑠(𝑠 + + 𝑠 + 1)

7/10/22 10:53 PM SK, Pulchowk Campus 74

 Assignment 3

7/10/22 10:53 PM SK, Pulchowk Campus 75

 Assignment 4

7/10/22 10:53 PM SK, Pulchowk Campus 76

 Assignment 5

7/10/22 10:53 PM SK, Pulchowk Campus 77