INSTITUTE OF AERONAUTICAL ENGINEERING

(Autonomous)
DUNDIGAL, HYDERABAD - 500043

CONTROL SYSTEMS
III B.Tech I semester (JNTUH-R13)

Prepared by
Dr. K Nehru
Professor
N Nagaraju
Assistant professor
ECE Department
Control Systems
Chapter 1: Introduction to Control Systems
Objectives
In this chapter we describe a general process for designing a control system.

A control system consisting of interconnected components is designed to achieve a

desired purpose. To understand the purpose of a control system, it is useful to
examine examples of control systems through the course of history. These early
systems incorporated many of the same ideas of feedback that are in use today.

Modern control engineering practice includes the use of control design strategies for
improving manufacturing processes, the efficiency of energy use, advanced
automobile control, including rapid transit, among others.

We also discuss the notion of a design gap. The gap exists between the complex
physical system under investigation and the model used in the control system
synthesis.

The iterative nature of design allows us to handle the design gap effectively while
accomplishing necessary tradeoffs in complexity, performance, and cost in order to
meet the design specifications.
Introduction

System – An interconnection of elements and devices for a desired purpose.

Control System – An interconnection of components forming a system configuration

that will provide a desired response.

Process – The device, plant, or system

under control. The input and output
relationship represents the cause-and-
effect relationship of the process.
Introduction
Open-Loop Control Systems
utilize a controller or control
actuator to obtain the desired
response.

Closed-Loop Control Systems

utilizes feedback to compare
the actual output to the
desired output response.

Multivariable Control System

History

Greece (BC) – Float regulator mechanism

Holland (16th Century)– Temperature regulator

Watt’s Flyball Governor

(18th century)
History

Water-level float regulator

History
History
18th Century James Watt’s centrifugal governor for the speed control of a steam
engine.
1920s Minorsky worked on automatic controllers for steering ships.
1930s Nyquist developed a method for analyzing the stability of controlled systems
1940s Frequency response methods made it possible to design linear closed-loop
control systems
1950s Root-locus method due to Evans was fully developed
1960s State space methods, optimal control, adaptive control and
1980s Learning controls are begun to investigated and developed.
Present and on-going research fields. Recent application of modern control theory
includes such non-engineering systems such as biological, biomedical, economic and
socio-economic systems
???????????????????????????????????
Examples of Modern Control Systems

(a) Automobile
steering control
system.
(b) The driver uses
the difference
between the actual
and the desired
direction of travel
to generate a
controlled adjustment
of the steering wheel.
(c) Typical direction-
of-travel response.
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
Examples of Modern Control Systems
The Future of Control Systems
The Future of Control Systems
Control System Design
Design Example
Design Example
ELECTRIC SHIP CONCEPT

Vision
Electrically
Reconfigurable
All
Integrated Ship
Electric
Power Ship
System  Technology
Insertion Increasing Affordability and Military Capability
 Reduced manning
 Electric Drive  Warfighting
 Automation
 Reduce # of Prime Capabilities
Movers  Eliminate auxiliary
systems (steam,
 Fuel savings hydraulics, compressed Main Power
 Reduced maintenance air) Distribution
Propulsion Motor Prime
Generator
Motor Drive Mover

Power Ship
Conversion Service
Module Power
Design Example

CVN(X) FUTURE AIRCRAFT CARRIER

Design Example
Design Example
Design Example
Design Example
Design Example
Design Example
Sequential Design Example
Sequential Design Example
Mathematical Models of Systems Objectives

We use quantitative mathematical models of physical systems to design and

analyze control systems. The dynamic behavior is generally described by
ordinary differential equations. We will consider a wide range of systems,
including mechanical, hydraulic, and electrical. Since most physical systems are
nonlinear, we will discuss linearization approximations, which allow us to use
Laplace transform methods.

We will then proceed to obtain the input–output relationship for components and
subsystems in the form of transfer functions. The transfer function blocks can be
organized into block diagrams or signal-flow graphs to graphically depict the
interconnections. Block diagrams (and signal-flow graphs) are very convenient
and natural tools for designing and analyzing complicated control systems
Introduction
Six Step Approach to Dynamic System Problems

Define the system and its components

Formulate the mathematical model and list the necessary
assumptions
Write the differential equations describing the model
Solve the equations for the desired output variables
Examine the solutions and the assumptions
If necessary, reanalyze or redesign the system
Differential Equation of Physical Systems

Ta( t)  Ts( t ) 0

Ta( t) Ts ( t )

( t ) s( t )  a( t)

Ta( t) = through - variable

angular rate difference = across-variable

Differential Equation of Physical Systems

Electrical Inductance Describing Equation Energy or Power

d 1 2
v21 L i E  L i
dt 2

Translational Spring
2
1 d 1 F
v 21  F E 
k dt 2 k

Rotational Spring
2
1 d 1 T
21  T E 
k dt 2 k

Fluid Inertia
d 1 2
P21 I Q E  I Q
dt 2
Differential Equation of Physical Systems
Electrical Capacitance
d 1 2
i C v 21 E  M v 21
dt 2
Translational Mass
d 1 2
F M v 2 E  M v 2
dt 2
Rotational Mass

d 1 2
T J 2 E  J 2
dt 2
Fluid Capacitance
d 1 2
Q Cf  P21 E  Cf  P21
dt 2
Thermal Capacitance

d
q Ct T2 E Ct T2
dt
Differential Equation of Physical Systems
Electrical Resistance
1 1 2
i  v21 P  v21
R R
Translational Damper

2
F b v21 P b v21

Rotational Damper
2
T b 21 P b 21

Fluid Resistance
1 1 2
Q  P21 P  P21
Rf Rf
Thermal Resistance
1 1
q  T21 P  T21
Rt Rt
Differential Equation of Physical Systems

2
d d
M y ( t )  b  y ( t )  k y ( t ) r( t )
2 dt
dt
Differential Equation of Physical Systems

t
v( t) d 1 
 C v( t)    v( t) dt r( t)
R dt L 0

  1 t
y( t) K 1 e  
 sin  1 t   1
Differential Equation of Physical Systems
Differential Equation of Physical Systems

K2  1  2  .5  2  10  2  2

  2 t
y ( t)  K2 e 
 sin  2 t   2 
  2 t   2 t
y1( t)  K2 e y2( t)  K2 e

y ( t)

y1( t ) 0
y2( t )

1
0 1 2 3 4 5 6 7
t
Linear Approximations
Linear Approximations

Linear Systems - Necessary condition

Principle of Superposition

Property of Homogeneity

Taylor Series
http://www.maths.abdn.ac.uk/%7Eigc/tch/ma2001/notes/node46.ht
ml
Linear Approximations – Example 2.1

m 15
M  200gm g  9.8 L  100cm  0  0rad      
s
2 16
 
T0  M  g  L sin  0

T1    M  g  L sin   

 
T2    M  g  L cos  0     0  T0 

10

T 1(  )
0
T 2(  )

10
4 3 2 1 0 1 2 3 4


0
Students are encourag ed to investigate linear approximation accuracy for different value
The Laplace Transform

Historical Perspective - Heaviside’s Operators

Origin of Operational Calculus (1887)

Historical Perspective - Heaviside’s Operators
Origin of Operational Calculus (1887)

t
d 1  v = H(t)
p  1 du
dt p 
0

i
v
Z( p ) R  L p Expanded in a power series
Z( p )

 R 1  R2 1  R3 1 
      .....  H( t)
1 1 1
i  H( t)  H( t)   
R  L p  L  L  p 3 
L p  1 
R  L p R  2
  p
 L p 

n
1 t
 H( t)
n n
p

 R 
   t
R  R t 2 2  3 3
1  L 
i
1
     R t
 t          .. 
 i 1  e   
R L  L  2  L  3   R

(*) Oliver Heaviside: Sage in Solitude, Paul J. Nahin, IEEE Press 1987.
The Laplace Transform

Definition


  s t
L( f ( t ) )  f ( t)  e dt = F(s)

0

Here the complex frequency is s   j w

The Laplace Transform exists when


  s t
 f ( t)  e dt  
 this means that the integral converges
0
The Laplace Transform

Determine the Lap lace transform for the functions

a) f1( t)  1 for t0


  s t
F1( s )   e dt 1  ( s t ) 1
 =  e
0 s s

 ( a t )
b) f2( t) e


  ( a t )  ( s t )
F2( s )  e e dt 1  [ ( s a)  t ] 1
 =  e F2( s )
0 s1 sa
The Laplace Transform
Evaluate the laplace transform of the derivative of a function



 d f ( t) 
L  d
f ( t)  e
 ( s t )
 
dt
 dt  
dt
0
 
by the use of  u dv = u v   v du
 
 ( s t )
where u e dv df ( t)

and, from which

 ( s t )
du s  e  dt
and v f ( t)
we obtain
 
 
f ( t)   s  e
 ( s t )  ( s  t )
 u dv = f ( t)  e   dt
 
0 0

  ( s t )
= -f(0+) + s f ( t)  e dt

0
L 
 d f ( t) 

 d t  = sF(s) - f(0+) note that the initial condition is included in the transformation
The Laplace Transform
Practical Example - Consider the circuit.

The K VL equation is

d
4 i( t)  2 i( t ) 0 assume i(0+) = 5 A
dt

Applying the Laplace Transform, we have

  
  
  4 i( t)  2 d i( t)   e  ( s t )
4 
 ( s t )
d t  2 
d  ( s t )

  dt 0

i( t )  e

i( t )  e dt 0
  dt  0 
dt
0 0

4 I( s )  2 ( s  I( s )  i( 0) ) 0 4 I( s )  2 s  I( s )  10 0

5 transforming back to the time domain, with our present knowledge of

I( s ) 
s2 Laplace transform, we may say that
t  ( 0  0.01 2)
6

 ( 2 t )
i( t)  5 e 4
i( t )
2

0
0 1 2
t
The Laplace Transform
The Partial-Fraction Expansion (or Heaviside expansion theorem)

Suppose that

s  z1 The partial fraction expansion indicates that F(s) consists of

F(s)
( s  p1 )  ( s  p2 ) a sum of terms, each of which is a factor of the denominator.
The values of K1 and K2 are determined by combining the
individual fractions by means of the lowest common
denominator and comparing the resultant numerator
or
coefficients with those of the coefficients of the numerator
K1 K2 before separation in different terms.
F(s) 
s  p1 s  p2

Evaluation of Ki in the manner just described requires the simultaneous solution of n equations.
An alternative method is to multiply both sides of the equation by (s + pi) then setting s= - pi, the
right-hand side is zero except for Ki so that

( s  pi )  ( s  z1 )
Ki s = - pi
( s  p1 )  ( s  p2 )
The Laplace Transform

Property Time Domain Frequency Domain

 ( s T )
e  F( s )
1. Time delay f ( t  T)  u ( t  T)

 F
1 s

2. Time scaling f ( at ) a  a

differentiation t f ( t) d
3. Frequency  F( s )
ds

 ( a t )
4. Frequency shifting f ( t)  e F( s  a)


f ( t) 
5. Frequency Integration  F( s ) d s
t 
0

6. Initial-value Theorem Lim( f ( t) ) f ( 0) Lim( s  F( s ) )

t -> 0 s -> infinite

7. Final-value Theorem Lim( f ( t) ) Lim( s  F( s ) )

t -> infinite s -> 0

The Laplace Transform

Useful Transform Pairs

 The Laplace Transform
Consider the mass-spring-damper system

( Ms  b )  yo
Y( s ) equation 2.21
2
Ms  bs  k

s  b  y
  
M o s  2  n
y(s )

s 2   b   s  k  2 2
    s  2  n  n
  M M

s1  
  n  n    1
2
k b
n 
M 2 k M

 
2
s2   n  n    1

Roots
s1  
  n  j n  1  
2

Real
Real repeated
 
2
Imaginary (conjug ates) s2   n  j n  1  
Complex (conjugates)
The Laplace Transform
The Transfer Function of Linear Systems

 R  1   I( s )
V1( s )   Z1( s ) R
 Cs 
1
 1 Z2( s )
V2( s )  Cs   I( s ) Cs
 
1
V2( s ) Cs Z2( s )
V1( s ) 1 Z1( s )  Z2( s )
R 
Cs
 The Transfer Function of Linear Systems

Example 2.2

d2 d The partial fraction expansion yields:

y( t)  4 y( t)  3 y ( t) 2 r( t)
2 dt
dt
 3 1  
 
1  2
2 2 1 3 3
d Y( s )   
    
Initial Conditions: Y( 0) 1 y ( 0) 0 r( t) 1  ( s  1) ( s  3)   ( s  1) ( s  3)  s
dt
The Laplace transform yields: Therefore the transient response is:

s 2Y(s )  s y(0)  4(s Y(s )  y(0))  3Y(s ) 2 R( s ) y ( t)  3  e t  1  e 3 t   1e t  1  e 3 t  2

2    3
 2   3 
Since R(s)=1/s and y(0)=1, we obtain:
The steady-state response is:
( s  4) 2

   
Y( s ) 2
2 2
s  4s  3 s  s  4s  3 lim y( t)
t  3
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems

 Kf if

Tm K1 Kf if( t )  ia( t )

fi el d control ed m otor - Lapal ce T ransform

Tm( s ) K1 Kf Ia If( s )
Vf( s )  Rf  Lf s  If( s )
Tm( s ) TL( s )  Td ( s )

2
TL( s ) J s   ( s )  b  s   ( s )

rearrangi ng equati ons

TL( s ) Tm( s )  Td ( s )

Tm( s ) Km If( s ) Td( s ) 0

Vf( s ) Km
If( s )  (s )
Rf  Lf s
Vf( s )  
s  ( J s  b )  Lf s  Rf
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems
The Transfer Function of Linear Systems

V 2( s ) 1
V 1( s ) RCs

V 2( s )
RCs
V 1( s )
The Transfer Function of Linear Systems

V 2( s ) R 2 R 1 C s  1
V 1( s ) R1

V 2( s ) R 1 C 1 s  1 R 2 C 2 s  1
V 1( s ) R 1 C 2 s
The Transfer Function of Linear Systems

 (s ) Km
V f(s ) s  ( J s  b)  L f  s  R f 

 (s ) Km
V a( s ) s  R a  L a s  ( J s  b)  K b K m
The Transfer Function of Linear Systems
 (s ) Km
Vc( s ) s   s  1

J

( b  m)

m = slope of linearized
torque-speed curve
(normally negative)

 K 
 R R 
Vo( s )  c q
Vc( s )  s  c  1   s  q  1
Lc Lq
c q
Rc Rq
For the unloaded case:
id 0 c q
0.05s  c  0.5s
V12 Vq V34 Vd
The Transfer Function of Linear Systems
Y( s ) K
X( s ) s ( Ms  B)
A  kx  A 
2
K B b  
kp  kp 

d d
kx g kp g
dx dP
g g ( x P) flow

A = area of piston

Gear Ratio = n = N1/N2

N2  L N1  m
L n  m

L n  m
The Transfer Function of Linear Systems

V2( s ) R2 R2
V1( s ) R R1  R2
R2 
R  max

V2( s ) 
ks  1( s )   2( s ) 
V2( s ) ks  error( s )

Vba ttery
ks
ma x
The Transfer Function of Linear Systems

V2( s ) Kt ( s ) Kt s   ( s )

Kt constant

V2( s ) ka
V1( s ) s  1

Ro = output resistance
Co = output capacitance
 Ro Co   1s
and is often negligible
for controller amplifier
The Transfer Function of Linear Systems
xo( t ) y ( t )  xin( t )
2
Xo( s ) s
Xin( s ) 2  b  s  k
s   
 M M

For low frequency oscillations, where  n

Xo j  
2

Xin j  k
M

T( s ) 1

Ct s   Q S 
q(s ) 1

 R
T To  Te = temperature difference due to thermal process
Ct = thermal capacitance
Q = fluid flow rate = constant
S = specific heat of water
Rt = thermal resistance of insulation
q( s ) = rate of heat flow of heating element
The Transfer Function of Linear Systems

x r 

converts radial motion to linear motion

Block Diagram Models
Block Diagram Models
Block Diagram Models
Original Diagram Equivalent Diagram

Original Diagram Equivalent Diagram

Block Diagram Models

Original Diagram Equivalent Diagram

Original Diagram Equivalent Diagram

Block Diagram Models
Original Diagram Equivalent Diagram

Original Diagram Equivalent Diagram

Block Diagram Models
Block Diagram Models
Example 2.7
Block Diagram Models Example 2.7
Signal-Flow Graph Models

For complex systems, the block diagram method can become

difficult to complete. By using the signal-flow graph model, the
reduction procedure (used in the block diagram method) is not
necessary to determine the relationship between system
variables.
Signal-Flow Graph Models

Y1( s ) G11( s )  R1( s )  G12( s )  R2( s )

Y2( s ) G21( s )  R1( s )  G22( s )  R2( s )

Signal-Flow Graph Models

a11 x1  a12 x2  r1 x1

a21 x1  a22 x2  r2 x2
Signal-Flow Graph Models
Example 2.8

Y( s ) G 1 G 2 G 3 G 4 1  L 3  L 4  G 5 G 6 G 7 G 8 1  L 1  L 2

R( s ) 1  L 1  L 2  L 3  L 4  L 1 L 3  L 1 L 4  L 2 L 3  L 2 L 4
Signal-Flow Graph Models
Example 2.10

Y( s ) G 1 G 2 G 3 G 4
R( s ) 1  G 2 G 3 H 2  G 3 G 4 H 1  G 1 G 2 G 3 G 4 H 3
Signal-Flow Graph Models

Y( s ) P1  P2  2  P3
R( s ) 

P1 G1 G2 G3 G4 G5 G6 P2 G1 G2 G7 G6 P3 G1 G2 G3 G4 G8
 1   L1  L2  L3  L4  L5  L6  L7  L8   L5 L7  L5 L4  L3 L4

1 3 1 2 1  L5 1  G4 H4
Design Examples
Design Examples

Speed control of an electric traction motor.

Design Examples
Design Examples
Design Examples
Design Examples
BLOCK DIAGRAM REDUCTION
OF MULTIPLE SYSTEMS
Components of a block diagram for a linear, time-invariant system
 a. Cascaded subsystems;
b. equivalent transfer function
 a. Parallel subsystems;
b. equivalent transfer function
 a. Feedback control system;
b. simplified model;
c. equivalent transfer function
Block diagram algebra for summing junctions

equivalent forms for moving a block

a. to the left past a summing junction;
b. to the right past a summing junction
 Block diagram algebra for pickoff points
equivalent forms for moving a block
a. to the left past a pickoff point;
b. to the right past a pickoff point
Block diagram reduction via familiar forms for Example

Problem: Reduce the block diagram shown in figure to a single transfer

function
Block diagram reduction via familiar forms for Example Cont.

Steps in solving Example

a. collapse summing junctions;
b. form equivalent cascaded
system in the forward path
c. form equivalent parallel
system in the feedback
path;
d. form equivalent feedback system and
multiply by cascadedG1(s)
Block diagram reduction by moving blocks Example

Problem: Reduce the block diagram shown in figure to a single transfer function
 Steps in the block diagram reduction for Example

a)Move G2(s) to the left past of

pickoff point to create parallel
subsystems, and reduce the
feedback system of G3(s) and H3(s)
b)Reduce parallel pair of 1/G2(s)
and unity, and push G1(s) to the
right past summing junction
c)Collapse the summing junctions,
add the 2 feedback elements, and
combine the last 2 cascade blocks
d)Reduce the feedback system to
the left
e) finally, Multiple the 2 cascade
blocks and obtain final result.
 Second-order feedback control system

K
The closed loop transfer function is T (s )  2
s  as  K
Note K is the amplifier gain, As K varies, the poles move through
the three ranges of operations OD, CD, and UD
0<K<a2/4 system is over damped
K = a2/4 system is critically damped
K > a2/4 system is under damped
 Finding transient response Example

Problem: For the system shown, find peak time, percent overshot, and settling time.
Solution: The closed loop transfer function is T (s )  25
s 2  5s  25
And n  2 5  5
2  n  5 s o  = 0 . 5
using values for  and n and equation in chapter 4 we find

Tp   0 .7 2 6 sec
n 1   2

1  2
%O S  e   / X 1 0 0  1 6 .3 0 3
4
Ts   1.6 sec
 n
 Gain design for transient response Example

Problem: Design the value of gain K, so that the system will respond with a 10%
overshot. K
T (s ) 
Solution: The closed loop transfer function is s 2  5s  K
5
 n  K and 2  n  5 th u s  =
2 K
For 10% OS we find  = 0 .5 9 1
We substitute this value in previous equation to find K = 17.9
 Signal-flow graph components:
a. system;
b. signal;
c. interconnection of systems and signals
Building signal-flow graphs

a. cascaded system
nodes
b. cascaded system
signal-flow graph;
c. parallel system nodes
d. parallel system
signal-flow graph;
e. feedback system nodes
f. feedback system signal-flow
graph
Converting a block diagram to a signal-flow graph

Problem: Convert the block diagram to a signal-flow graph.

 Converting a block diagram to a signal-flow graph
Signal-flow graph development:
a. signal nodes;
b. signal-flow graph;
c. simplified signal-flow graph
 Mason’s ƌule - Definitions

Loop gain: The product of branch gains found by traversing a path that starts at a node and ends at the
same node, following the direction of the signal flow, without passing through any other node more than
once. G2(s)H2(s), G4(s)H2(s), G4(s)G5(s)H3(s), G4(s)G6(s)H3(s)
Forward-path gain: The product of gains found by traversing a path from input node to output node in
the direction of signal flow. G1(s)G2(s)G3(s)G4(s)G5(s)G7(s), G1(s)G2(s)G3(s)G4(s)G5(s)G7(s)
Nontouching loops: loops that do not have any nodes in common. G2(s)H1(s) does not touch G4(s)H2(s),
G4(s)G5(s)H3(s), and G4(s)G6(s)H3(s)
Nontouching-loop gain: The product of loop gains from nontouching loops taken 2, 3,4, or more at a
time.
[G2(s)H1(s)][G4(s)H2(s)], [G2(s)H1(s)][G4(s)G5(s)H3(s)], [G2(s)H1(s)][G4(s)G6(s)H3(s)]
 Mason’s Rule
The Transfer function. C(s)/ R(s), of a system represented by a signal-flow graph is

G (s ) 
C (s )

T k k
k

Where R (s ) 
K = number of forward paths
Tk = the kth forward-path gain

= 1 - loop gains + nontouching-loop gains taken 2 at a time - nontouching-loop

  
gains taken 3 at a time + nontouching-loop gains taken 4 at a 
time - …….

= - loop gain terms in that touch the kth forward path. In other words,
k is for y eliminating from  those loop gains that touch the kth forward path. k
med b

 Transfer function via Mason’s rule
Problem: Find the transfer function for the signal flow graph
Solution:
forward path
G1(s)G2(s)G3(s)G4(s)G5(s)
Loop gains
G2(s)H1(s), G4(s)H2(s), G7(s)H4(s),
G2(s)G3(s)G4(s)G5(s)G6(s)G7(s)G8(s)
Nontouching loops
2at a time
G2(s)H1(s)G4(s)H2(s)
G2(s)H1(s)G7(s)H4(s)
G4(s)H2(s)G7(s)H4(s)
3at a time G2(s)H1(s)G4(s)H2(s)G7(s)H4(s)
Now
 = 1-[G2(s)H1(s)+G4(s)H2(s)+G7(s)H4(s)+ G2(s)G3(s)G4(s)G5(s)G6(s)G7(s)G8(s)] + [G2(s)H1(s)G4(s)H2(s) +
G2(s)H1(s)G7(s)H4(s) + G4(s)H2(s)G7(s)H4(s)] – [G2(s)H1(s)G4(s)H2(s)G7(s)H4(s)]
1 = 1 - G7(s)H4(s)
T 1 1
G (s )  [G (s)G2(s)G3(s)G4(s)G5(s)][1-G7(s)H4(s)]
1
 
Signal-Flow Graphs of State Equations
Problem: draw signal-flow graph for:
x 1  2 x 1  5x 2  3x 3  2 r

x 2   6 x 1  2 x 2  2 x 3  5r
x 3  x 1  3x 2  4 x 3  7 r
y  4x 1  6x 2  9x 3
a. place nodes;
b. interconnect state
variables and
derivatives;
c. form dx1/dt ;
d. form dx2/dt
Signal-Flow Graphs of State Equations

(continued)
e. form dx3 /dt;
f. form output
 Alternate Representation: Cascade Form

C (s ) 24

R (s ) (s  2)(s  3)(s  4)
 Alternate Representation: Cascade Form

 4 1 0  0 
x 1  4x 1  x 2
X   0 3 1  X   0  r
x2   3x 2  x 3    
 0 2   24 
x3   2x 3  24r  0 
y  c (t )  x 1 0X
y  1 0
 C (s ) Alternat2e4Representation1: Parallel
2 F2o4rm 12
   
R (s ) (s  2)(s  3)(s  4) (s  2) (s  3) (s  4)

x 1  2 x 1  12r
x2   3x 2  24r
x3   4x 3  12r
y  c (t )  x 1  x 2  x 3

 2 0 0  12 
X   0 3 0  X   24  r
   
 0 0 4  12 
y  1 1 1X
 Alternate Representation: Parallel Form Repeated roots
C (s ) (s  3) 2 1 1
 
R (s ) (s  1) (s  2) (s  1) 2
2
(s  1) (s  2)

x 1  x 1  x 2
x2  x2 + 2r
x3   2x 3  r
y  c (t )  x 1  1 / 2 x 2  x 3

 1 1 0  0 
X   0 1 0  X   2  r
   
 0 0 2  1 
y  1 1 / 2 1X
 Alternate Representation: controller canonical form

G(s) = C(s)/R(s) = (s2 + 7s + 2)/(s3 + 9s2 + 26s + 24)

This form is obtained from the phase-variable form simply by ordering the
phase variable in reverse order

x1  x1 
   
 0 1 0   x 1   0  9 26 24  x 1  1 
             
x 2    0 0 1   x 2    0 r x 2    1 0 0   x 2    0 r
   24 26 9  x  1     0 0   x 3  0
  3   x 3 
1
x 3 
   
x 1  x 1 
   
y  2 7 1 x 2  y  1 7 2 x 2 
 x 3 
 x 3 
 Alternate Representation: controller canonical form

System matrices that contain the coefficients of the characteristic polynomial are
called companion matrices to the characteristic polynomial.
Phase-variable form result in lower companion matrix
Controller canonical form results in upper companion matrix
 Alternate Representation: observer canonical form

Observer canonical form so named for its use in the design of observers
G(s) = C(s)/R(s) = (s2 + 7s + 2)/(s3 + 9s2 + 26s + 24)
= (1/s+7/s2 +2/s3 )/(1+9/s+26/s2 +24/s3 )
Cross multiplying
(1/s+7/s2 +2/s3 )R(s) = (1+9/s+26/s2 +24/s3 ) C(s)
And C(s) = 1/s[R(s)-9C(s)] +1/s2[7R(s)-26C(s)]+1/s3[2R(s)-24C(s)]
= 1/s{ [R(s)-9C(s)] + 1/s {[7R(s)-26C(s)]+1/s [2R(s)-24C(s)]}}
 Alternate Representation: observer canonical form

x 1  9x 1  x 2 r
x 2  26x 1  x 3 +7r

x 3  24x 1  2r
y  c (t )  x 1

 9 1 0 1 
X   26 0 1  X  7  r
   
 24 0 0   2 
y  1 0 0X
Note that the observer form has A matrix that is transpose of the
controller canonical form, B vector is the transpose of the controller C
vector, and C vector is the transpose of the controller B vector. The 2
forms are called duals.
 Feedback control system for Example

Problem Represent the feedback control system

shown in state space. Model the forward transfer
function in cascade form.
Solution first we model the forward transfer
function as in (a), Second we add the feedback
and input paths as shown in (b) complete system.
Write state equations

x 1  3x 1  x 2
x2  - 2 x 2  1 00 (r - c )

but c  5x 1  (x 2  3x 1 )  2 x 1  x 2
 Feedback control system for Example

x 1  3x 1  x 2
x 2  200 x 1  102 x 2  100 r
y  c (t )  2 x 1  x 2

 3 1  X  0 
X    r
 200 102  100 
y  2 1X
 State-space forms for

C(s)/R(s) =(s+ 3)/[(s+ 4)(s+ 6)].

Note: y = c(t)
 UNIT-II
TIME RESPONSE ANALYSIS
Feedback Control System Characteristics
Objectives

In this chapter we extend the ideas of modeling to include control system

characteristics, such as sensitivity to model uncertainties, steady-state errors,
transient response characteristics to input test signals, and disturbance
rejection. We investigate the important role of the system error signal which we
generally try to minimize.

We will also develop the concept of the sensitivity of a system to a parameter

change, since it is desirable to minimize the effects of unwanted parameter
variation. We then describe the transient performance of a feedback system
and show how this performance can be readily improved. We will also
investigate a design that reduces the impact of disturbance signals.
Open-And Closed-Loop Control Systems

An open-loop (direct) system

operates without feedback and
directly generates the output in
response to an input signal.

A closed-loop system uses a

measurement of the output
signal and a comparison with
the desired output to generate
an error signal that is applied to
the actuator.
Open-And Closed-Loop Control Systems

H( s ) 1

G( s )
Y( s )  R( s )
1  G( s )

1
E( s )  R( s ) Error Signal
1  G( s )

Thus, to reduce the error, the magnitude of 1  G( s )   1

H( s )  1

G( s )
Y( s )  R( s )
1  H( s ) G( s )

1
E( s )  R( s )
1  H[ ( s )  G( s ) ]

Thus, to reduce the error, the magnitude of 1  G( s ) H( s )   1

Sensitivity of Control Systems To Parameter Variations

For the closed-loop case if GH( s )  1

1
Y ( s) R ( s) Output affected only by H(s)
H(s)

G( s )  G( s )

Open Loop Y ( s ) G( s )  R ( s )

Closed Loop

G( s )  G( s ) 
Y ( s )  Y ( s ) R ( s )
1  G( s )  G( s )  H ( s )

G( s )
Y ( s ) R ( s)
1  GH( s )  GH( s ) ( 1  GH( s ) )

GH( s )  GH ( s )

G( s ) The change in the output of the closed system

Y ( s ) R ( s )
2 is reduced by a factor of 1+GH(s)
( 1  GH( s ) )
 Sensitivity of Control Systems To Parameter Variations

Y ( s )
T ( s )
R ( s )
d
T
 T ( s) d T  d T
 
S
T ( s)
S
T d T  
G
 G ( s)  d G  d G T
d G   
G ( s)
  d G 
 G 
1
T ( s )
1  H [ ( s ) G ( s ) ]

 d T  d T
   
T d T  
G d T  
G 1

G
S G
 d  T  d  T
( 1 
2 G
 
G  G GH )
d G  d G  ( 1  GH )

T 1
S G Sensitivity of the closed-loop to G variations reduced
( 1  GH )

T  GH Sensitivity of the closed-loop to H variations

S H When GH is large sensitivity approaches 1
( 1  GH )
Changes in H directly affects the output response
Example 4.1

Open loop Closed loop

vo Ka v in R2
 Rp R1  R2
R1
T ka Ka T 1
T SKa
T 1  Ka  1  Ka 
SKa 1

If Ka is large, the sensitivity is low.

4 T 1 4
Ka  10   0.1 SKa   9.99  10
3
1  10
Control of the Transient Response of Control Systems

( s ) K1
G( s )
Va( s )  1 s  1

where,

Km Ra J
K1 1
Ra b  Kb Km Ra b  Kb Km
Control of the Transient Response of Control Systems

K1
K a
( s ) K a G( s ) K a K 1 1
R( s ) 1  K a K t G( s )  1 s  1  K a K t K 1   1  K a K t K 1
s   
  1 
Control of the Transient Response of Control Systems
 Disturbance Signals In a Feedback Control Systems

R(s)
Disturbance Signals In a Feedback Control Systems
 Disturbance Signals In a Feedback Control Systems

Km 1 Kb
G1( s ) Ka G2( s ) H( s ) Kt 
Ra ( J s  b ) Ka

G( s )
E( s ) ( s )  Td( s )
1  G1( s )  G2( s )  H( s )

( )1
G1G2Hs

1
E( s )  Td( s ) If G1(s)H(s) very large the effect of the disturbance
G1( s )  H( s ) can be minimized

Ka Km  Kb  Ka Km Kt

G1( s ) H( s )  Kt   approximately since Ka >> Kb
Ra  Ka  Ra

Strive to maintain Ka large and Ra < 2 ohms

Steady-State Error

Eo ( s ) R( s )  Y( s ) ( 1  G( s ) )  R( s )

1
Ec( s )  R( s ) H( s ) 1
1  G( s )

Steady State E rror

lim e( t ) lim s  E( s )
t 0 s 0

For a step unit input

1
eo ( infinite) lim s  ( 1  G( s ) )  lim ( 1  G( 0) )
s 0 s s 0

s  1  
1 1
ec( infinite) lim   lim  
s 0  1  G( s )  s s 0  1  G( 0) 
The Cost of Feedback

Increased Number of components and Complexity

Loss of Gain

Instability
Design Example: English Channel Boring Machines

Y( s ) T( s )  R( s )  Td( s )  D( s )

K  11 s 1
Y( s )  R( s )   D( s )
2 2
s  12 s  K s  12 s  K
Design Example: English Channel Boring Machines

Study system for different

Values of gain K

Steady state error for R(s)=1/s and D(s)=0

lim s     1
1
lim e( t ) 0
t  infinite s 0  1  K  11 s  s
 G
s  s 
2

Steady state error for R(s)=0 D(s)=1/s

lim s    1
1 1
lim
t  infinite
y( t)
s 0
 2 Td  s
 s  12 s  K 
K
 Transient vs Steady-State
The output of any differential equation can be broken up into two parts,
•a transient part (which decays to zero as t goes to infinity) and
•a steady-state part (which does not decay to zero as t goes to infinity).

y(t)  ytr (t)  yss (t)

lim ytr (t)  0
t0

Either part might be zero in any particular case.

 Prototype systems
1
1st Order system c(t)  c(t)  kr(t)

2nd order system c(t)  2 nc(t)   n2 c(t)  kr(t)
Agenda:
transfer function
response to test signals
impulse
step ramp
parabolic
sinusoidal
1st order system G(s) 
C(s)

1T
Impulse response
Step response R(s) s 1 T
Ramp response
Relationship between impulse, step and ramp
Relationship between impulse, step and ramp responses

1
c (t)  e t T 1(t)
r(t)   (t), R(s)  1, 
1 T
r(t)  1(t), R(s)  ,
c (t)  1 e t T
 1(t)
s step  
1
r(t)  t1(t), R(s)  ,
cramp (t)  t T  Te t T  1(t)
s2
 1st Order system
Prototype parameter: Time constant

Relate problem specific parameter to prototype parameter.

Parameters: problem specific constants. Numbers that do not change with

time, but do change from problem to problem.

We learn that the time constant defines a problem specific time scale that is more
convenient than the arbitrary time scale of seconds, minutes, hours, days, etc, or
fractions thereof.
 Transient vs Steady state
Consider the impulse, step, ramp responses computed earlier. Identify the steady
state and the transient parts.
1st order Consider the impulse, step, ramp responses computed
earlier. Identify the steady state and the transient
system parts.
Impulse response C(s)
Step response G(s)   1T
Ramp response , T 0
Relationship between impulse, step and ramp R(s) s 1 T
Relationship between impulse, step and ramp responses

1
c (t)  e t T 1(t)
r(t)   (t), R(s)  1, 
1 T
r(t)  1(t), R(s)  ,
c (t)  1 e t T
 1(t)
s step  
1
r(t)  t1(t), R(s)  ,
cramp (t)  t T  Te t T  1(t)
s2
Compare steady-state part to input function, transient part to TF.
 C(s) K
2
2nd order system G(s)   n

R(s) s 2  2 s   2
n n
Over damped
•(two real distinct roots = two 1st order systems with real poles)
Critically damped
•(a single pole of multiplicity two, highly unlikely, requires exact matching)
Underdamped
•(complex conjugate pair of poles, oscillatory behavior, most common)
step response

 n t 
cstep (t)  K 1 e
 1  2
sin d t  tan
1
 1  2
  1(t)

 
  
sin d t  1(t)
nt
c (t)  K  n
e
 1  2 
 2nd Order System

Prototype parameters:
undamped natural frequency,
damping ratio

Relating problem specific parameters to prototype parameters

 Transient vs Steady state
Consider the step, responses computed earlier. Identify the steady state and the
transient parts.
 C(s) K
2
2nd order system G(s)   n

R(s) s 2  2 s   2
n n
Over damped
•(two real distinct roots = two 1st order systems with real poles)
Critically damped
•(a single pole of multiplicity two, highly unlikely, requires exact matching)
Underdamped
•(complex conjugate pair of poles, oscillatory behavior, most common)
step response

 n t 
cstep (t)  K 1 e
 1  2
sin d t  tan
1
 1  2
  1(t)

 
  
sin d t  1(t)
nt
c (t)  K  n
e
 1  2 
 Use of Prototypes

Too many examples to cover them all

We cover important prototypes
We develop intuition on the prototypes
We cover how to convert specific examples to prototypes
We transfer our insight, based on the study of the prototypes to the specific
situations.
 Transient-Response Spedifications
1. Delay time, td: The time required for the response to reach half the final value the
very first time.
2. Rise time, tr: the time required for the response to rise from
10% to 90% (common for overdamped and 1st order systems);
5% to 95%;
or 0% to 100% (common for underdamped systems);
of its final value
1. Peak time, tp:
2. Maximum (percent) overshoot, Mp:
3. Settling time, ts
 Derived relations ford2ndn 1  2
Order Systems   n
    d  1
tr  tp    tan  
d d  

 See book for details. (Pg. 232)
1 2
Mp e 100%
4 3
t  4T   4 t  3T   3
2% s 5%
  n  n
s

Allowable Mp determines damping ratio.

Settling time then determines undamped natural frequency.
Theory is used to derive relationships between design specifications and prototype
parameters.
Which are related to problem parameters.
 Higher order system

PFEs have linear denominators.

•each term with a real pole has a time constant

•each complex conjugate pair of poles has a damping ratio and an undamped
natural frequency.
Proportional control of plant w
integrator

1
GC (s)  K p , G(s) 
s(Js  b)
Integral control of Plant w disturbance

K
G (s)  , 1
G(s) 
s(Js  b)
C
s
Proportional Control of plant w/o
integrator

1
GC (s)  K, G(s) 
Ts 1
Integral control of plant w/o integrator

K
G (s)  , 1
G(s) 
Ts 1
C
s
 UNIT-III
STABILITY ANALYSIS IN S- DOMAIN
Chapter 6 – The Stability of Linear Feedback Systems

The issue of ensuring the stability of a closed-loop feedback system is central to

control system design. Knowing that an unstable closed-loop system is generally
of no practical value, we seek methods to help us analyze and design stable
systems. A stable system should exhibit a bounded output if the corresponding
input is bounded. This is known as bounded-input, bounded-output stability
and is one of the main topics of this chapter.

The stability of a feedback system is directly related to the location of the roots
of the characteristic equation of the system transfer function. The Routh–
Hurwitz method is introduced as a useful tool for assessing system stability. The
technique allows us to compute the number of roots of the characteristic
equation in the right half-plane without actually computing the values of the
roots. Thus we can determine stability without the added computational
burden of determining characteristic root locations. This gives us a design
method for determining values of certain system parameters that will lead to
closed-loop stability. For stable systems we will introduce the notion of relative
stability, which allows us to characterize the degree of stability.
The Concept of Stability

A stable system is a dynamic system with a bounded

response to a bounded input.

Absolute stability is a stable/not stable characterization for a

closed-loop feedback system. Given that a system is stable
we can further characterize the degree of stability, or the
relative stability.
 The Concept of Stability
The concept of stability can be
illustrated by a cone placed on
a plane horizontal surface.

A necessary and
sufficient condition for a
feedback system to be
stable is that all the
poles of the system
transfer function have
negative real parts.

A system is considered marginally stable if only certain bounded

inputs will result in a bounded output.
 The Routh-Hurwitz Stability Criterion

It was discovered that all coefficients of the characteristic polynomial must

have the same sign and non-zero if all the roots are in the left-hand plane.

These requirements are necessary but not sufficient. If the above

requirements are not met, it is known that the system is unstable. But, if the
requirements are met, we still must investigate the system further to
determine the stability of the system.

The Routh-Hurwitz criterion is a necessary and sufficient criterion for the

stability of linear systems.
 The Routh-Hurwitz Stability Criterion
Characteristic equation, q(s) an s n  an 1s n 1  an  2 s n  2    a1s  a0  0

Routh array
sn an an  2 an  4
s n 1 an 1 an 3 an 5
s n2 bn 1 bn 3 bn 5
s n 3 cn 1 cn  3 cn  5
   
The Routh-Hurwitz criterion    
states that the number of    
roots of q(s) with positive real
s0 hn 1
parts is equal to the number
of changes in sign of the first
column of the Routh array. bn 1 
an1 an2   an an3    1 an an  2
an 1 an 1 an 1 an  3
 1 an  2 an  4
bn 3 
an 1 an 1 an 3
 1 an 1 an 3
cn 1 
bn 1 bn 1 bn 3
The Routh-Hurwitz Stability Criterion
Case One: No element in the first column is zero.

Example 6.1 Second-order system

The Characteristic polynomial of a second-order sys tem is:

2
q(s ) a 2 s  a 1 s  a 0

The Routh array is w ritten as: 2

s a2 a0
1
s a1 0
0
w here: s b1 0
a1 a0  ( 0)  a2
b1 a0
a1

Theref ore the requirement f or a stable second-order system is

simply that all coef f icients be positive or all the coef ficients be
negative.
The Routh-Hurwitz Stability Criterion
Case Two: Zeros in the first column while some elements of the row containing a
zero in the first column are nonzero.

If only one element in the array is z ero, it may be replaced w ith a small positiv e
number  that is allow ed to approach zero after completing the array.

5 4 3 2
q( s ) s  2s  2s  4s  11s  10

The Routh array is then:

s 5 1 2 11
s 4 2 4 10
s 3 b1 6 0
s 2 c1 10 0
s1 d1 0 0
s 0 10 0 0
w here:
2 2  1 4 4  2 6 12 6 c1  10
b1 0  c1 d1 6
2   c1
There are tw o sign changes in the first column due to the large negative number
calculated f or c1. Thus, the system is unstable because tw o roots lie in the
right half of the plane.
The Routh-Hurwitz Stability Criterion
Case Three: Zeros in the first column, and the other elements of the row containing
the zero are also zero.

This case occurs w hen the polynomial q(s) has zeros located sy metrically about the
origin of the s-plane, such as (s+)(s -) or (s+j)(s -j). This c ase is solved using
the auxiliary poly nomial, U(s), w hich is located in the row above the row containing
the zero entry in the Routh array.

3 2
q( s ) s  2 s  4s  K

Routh array: s3 1 4
s2 2 K
8 K
s1 2 0
s0 K 0
For a stable system w e require that 0s 8

For the marginally stable case, K=8, the s^1 row of the Routh array contains all zeros. The
auxiliary plynomial comes f rom the s^2 row .
U( s )
2
2s  Ks
0 2
2 s  8 2 
2s 4 2( s  j 2) ( s  j 2)

It c an be proven that U(s) is a f actor of the characteris tic polynomial:

q( s ) s2
U( s ) 2 Thus, w hen K=8, the factors of the characteristic polynomial are:

q( s ) ( s  2) ( s  j 2) ( s  j 2)
The Routh-Hurwitz Stability Criterion
Case Four: Repeated roots of the characteristic equation on the jw-axis.

With simple roots on the jw-axis, the system will

have a marginally stable behavior. This is not
the case if the roots are repeated. Repeated
roots on the jw-axis will cause the system to be
unstable. Unfortunately, the routh-array will fail
to reveal this instability.
Example 6.4
Example 6.5 Welding control

Using block diagram reduction we find that:

The Routh array is then: s4 1 11 Ka
s3 6 ( K  6)
s2 b3 Ka
s1 c3
s0 Ka
60  K b3( K  6)  6 Ka
where: b3 and c3
6 b3
For the system to be stable bothb3 and c3 must be positive.
Using these equations a relationship can be determined for K and a .
The Relative Stability of Feedback Control Systems

It is often necessary to know the

relative damping of each root to
the characteristic equation.
Relative system stability can be
measured by observing the
relative real part of each root. In
this diagram r2 is relatively more
stable than the pair of roots
labeled r1.

One method of determining the relative stability of

each root is to use an axis shift in the s-domain and
then use the Routh array as shown in Example 6.6
of the text.
Design Example: Tracked Vehicle Turning Control

Problem statement: Design the turning control for a tracked vehicle. Select K
and a so that the system is stable. The system is modeled below.
Design Example: Tracked Vehicle Turning Control

The characteristic equation of this system is:

1  Gc G( s ) 0

or
K( s  a)
1 0
s ( s  1) ( s  2) ( s  5)

Thus,
s ( s  1) ( s  2) ( s  5)  K( s  a) 0

or
4 3 2
s  8s  17s  ( K  10)s  Ka 0
To determine a stable region for the system, we establish the Routh array as:
s4 1 17 Ka
s3 8 ( K  10) 0
s2 b3 Ka
s1 c3
s0 Ka
where
126  K b 3( K  10)  8Ka
b3 and c3
8 b3
Design Example: Tracked Vehicle Turning Control

s4 1 17 Ka
s3 8 ( K  10) 0
s2 b3 Ka
s1 c3
s0 Ka

where
126  K b 3( K  10)  8Ka
b3 and c3
8 b3

Therefore,
K  126

K a  0

( K  10) ( 126  K)  64Ka  0

 The Root Locus Method
In the preceding chapters we discussed how the performance of a feedback system can be
described in terms of the location of the roots of the characteristic equation in the s-plane. We
know that the response of a closed-loop feedback system can be adjusted to achieve the
desired performance by judicious selection of one or more system parameters. It is very useful
to determine how the roots of the characteristic equation move around the s-plane as we
change one parameter.

The locus of roots in the s-plane can be determined by a graphical method. A graph of the locus
of roots as one system parameter varies is known as a root locus plot. The root locus is a
powerful tool for designing and analyzing feedback control systems and is the main topic of this
chapter. We will discuss practical techniques for obtaining a sketch of a root locus plot by hand.
We also consider computer-generated root locus plots and illustrate their effectiveness in the
design process. The popular PID controller is introduced as a practical controller structure.

We will show that it is possible to use root locus methods for design when two or three
parameters vary. This provides us with the opportunity to design feedback systems with two or
three adjustable parameters. For example the PID controller has three adjustable parameters.
We will also define a measure of sensitivity of a specified root to a small incremental change in
a system parameter.
The Root Locus Method
The Root Locus Method

The root locus is a graphical procedure for determining the poles of a closed-loop system
given the poles and zeros of a forward-loop system. Graphically, the locus is the set of
paths in the complex plane traced by the closed-loop poles as the root locus gain is varied
from zero to infinity.

In mathematical terms, given a forward-loop transfer function, KG(s)

where K is the root locus gain, and the corresponding closed-loop transfer function

the root locus is the set of paths traced by the roots of

as K varies from zero to infinity. As K changes, the solution to this equation changes.
This equation is called the characteristic equation. This equation defines where the poles
will be located for any value of the root locus gain, K. In other words, it defines the
characteristics of the system behavior for various values of controller gain.
The Root Locus Method
The Root Locus Method
The Root Locus Method
The Root Locus Method
The Root Locus Method

No matter what we pick K to be, the closed-loop system must always have n poles, where
n is the number of poles of G(s).
The root locus must have n branches, each branch starts at a pole of G(s) and goes to a
zero of G(s).
If G(s) has more poles than zeros (as is often the case), m < n and we say that G(s) has
zeros at infinity. In this case, the limit of G(s) as s -> infinity is zero.

The number of zeros at infinity is n-m, the number of poles minus the number of zeros,
and is the number of branches of the root locus that go to infinity (asymptotes).

Since the root locus is actually the locations of all possible closed loop poles, from the
root locus we can select a gain such that our closed-loop system will perform the way we
want. If any of the selected poles are on the right half plane, the closed-loop system will
be unstable. The poles that are closest to the imaginary axis have the greatest influence on
the closed-loop response, so even though the system has three or four poles, it may still
act like a second or even first order system depending on the location(s) of the dominant
pole(s).
Example
Example

MATLAB Example - Plotting the root locus of a transfer function

Consider an open loop system which has a transfer function of
G(s) = (s+7)/s(s+5)(s+15)(s+20)
How do we design a feedback controller for the system by using the root locus method?
Enter the transfer function, and the command to plot the root locus:
num=[1 7];
den=conv(conv([1 0],[1 5]),conv([1 15],[1 20]));
rlocus(num,den)
axis([-22 3 -15 15])
Graphical
Method
Graphical
Method
Root Locus Design GUI (rltool)

The Root Locus Design GUI is an interactive

graphical tool to design compensators using
the root locus method. This GUI plots the
locus of the closed-loop poles as a function of
the compensator gains. You can use this GUI
to add compensator poles and zeros and
analyze how their location affects the root
locus and various time and frequency domain
responses. Click on the various controls on
the GUI to see what they do.
 UNIT-IV
FREQUENCY RESPONSE ANALYSIS
Frequency Response Methods
and Stability
In previous chapters we examined the use of test signals such as a step and a ramp
signal. In this chapter we consider the steady-state response of a system to a sinusoidal
input test signal. We will see that the response of a linear constant coefficient system
to a sinusoidal input signal is an output sinusoidal signal at the same frequency as the
input. However, the magnitude and phase of the output signal differ from those of the
input sinusoidal signal, and the amount of difference is a function of the input
frequency. Thus we will be investigating the steady-state response of the system to a
sinusoidal input as the frequency varies.

We will examine the transfer function G(s) when s =jw and develop methods for
graphically displaying the complex number G(j)as w varies. The Bode plot is one of the
most powerful graphical tools for analyzing and designing control systems, and we will
cover that subject in this chapter. We will also consider polar plots and log magnitude
and phase diagrams. We will develop several time-domain performance measures in
terms of the frequency response of the system as well as introduce the concept of
system bandwidth.
Introduction

The frequency response of a system is defined as the steady-state response of the

system to a sinusoidal input signal. The sinusoid is a unique input signal, and the
resulting output signal for a linear system, as well as signals throughout the system, is
sinusoidal in the steady-state; it differs form the input waveform only in amplitude and
phase.
Frequency Response Plots

Polar Plots
Frequency Response Plots

Polar Plots

1
  1000 999 1000 j  1 R  1 C  0.01 1 
R C

G  
1
Negative
 j   1

 
 1  0.5

 

Im( G(  ) ) 0

 0
 0
 
0.5
0 0.5 1
Re( G(  ) )
 1

Positive 
Frequency Response Plots

Polar Plots

7
0
 410
  0  .1 1000   0.5 K  100

K

G1   Im( G1(  ) ) 500

j   j  
1

 

 997.5061000
60 40 20 0
 49.875 Re( G1(  ) ) 4
 210
Frequency Response Plots

Polar Plots
Frequency Response Plots

Bode Plots – Real Poles

Frequency Response Plots

Bode Plots – Real Poles

0.1 0.11
    1000 j  1 R  1 C  0.01   R C
 

1
1  1  100 (break frequency or corner frequency)
G  
1

j    1

0
-3dB

10

20 log G(  ) 
20

30
1 10
3
0.1 1 10 100

(break frequency or corner frequency)
Frequency Response Plots

Bode Plots – Real Poles

   atan   

0.5

(  )

1.5
1 10
3
0.1 1 10 100


(break frequency or corner frequency)

Frequency Response Plots

Bode Plots – Real Poles (Graphical Construction)

Frequency Response Plots

Bode Plots – Real Poles

Frequency Response Plots

Bode Plots – Real Poles

 Frequency Response Plots

M agnitude: Bode Plots – Real Poles

db  G     20 log G  j   
Phase shift:

ps G     arg G  j     360  if  arg G  j     0  1  0 

180

Assume
K
K  2 G ( s) 

s ( 1  s)  1 
s

 3

Next, choose a frequency rangefor the plots (use powers of 10 for convenient plotting):

lowest frequency (in Hz):  start  .01 number of p oints: N  50

highest frequency (in Hz):  end  100

  start  1
step size: r  log 

 end  N
range for plot: i  0  N range variable: i r
 i   end 10 si  j  i
Frequency Response Plots

Bode Plots – Real Poles

range for plot: i  0  N range variable: i r

 i   end10 si  j  i

100

20 log G  s i  0

0
100

200
0.01 0.1 1 10 100
i

ps  G  i

 180

0.01 0.1 1 10 100

i
Frequency Response Plots

Bode Plots – Complex Poles

Frequency Response Plots

Bode Plots – Complex Poles

Frequency Response Plots

Bode Plots – Complex Poles

2
r n  1  2    0.707

Mp  
G r

1
2   0.707
 2  1   
 Frequency Response Plots

Bode Plots – Complex Poles

2
r n  1  2    0.707

 
1
M p G r
 2   0.707
 2  1   
Frequency Response Plots

Bode Plots – Complex Poles

Frequency Response Plots

Bode Plots – Complex Poles

Performance Specification In the Frequency Domain
Performance Specification In the Frequency Domain

  .1 .11 2 K  2 j  1

G  
K
j   j   1   j   2 Bode1   20 log G  
20

Open Loop Bode Di agram

Bo de1(  ) 0

20
0 0 .5 1 1 .5 2

G 

T  
1  G  Bode2   20 log T  

10

0
Bo de2(  )

10 Cl osed-Loop Bode Di agram

20
0 .5 1 1 .5 2

Performance Specification In the Frequency Domain

w  4
Finding the Resonance Frequency
Given

20 log T( w)  5.282

wr  Find( w) wr  0.813

Mpw  1

Given

20 log( Mpw) 5.282 Finding Maximum value of the frequency response

Mpw  Find( Mpw) Mpw  1.837 10

0
Bode2(  )

10 Closed-Loop Bode Diagram

20
0.5 1 1.5 2

Performance Specification In the Frequency Domain

Assume that the system has dominant second-order roots

  .1
Finding the damping factor
Given 10

1
  2
Mpw 2   1    0
Bode2(  )

  Find 
10 Closed-Loop Bode Diagram
  0.284
20
wn  .1 0.5 1 1.5 2
Finding the natural frequency 

Given
2
wr wn  1  2 

wn  Find( wn ) wn  0.888
Performance Specification In the Frequency Domain
 Performance Specification
In the Frequency Domain

GH1 
5

  
j  0.5 j   1   j
 
 1
 6 
Performance Specification In the Frequency Domain
Example
Performance Specification In the Frequency Domain
Example
Performance Specification In the Frequency Domain
Example
Performance Specification In the Frequency Domain
Example
Frequency Response Methods Using MATLAB
Frequency Response Methods Using MATLAB
Frequency Response Methods Using MATLAB
Frequency Response Methods Using MATLAB
Frequency Response Methods Using MATLAB
Frequency Response
Methods Using
MATLAB
Frequency Response
Methods Using
MATLAB
Bode Plots

Bode plot is the representation of the magnitude and phase of G(j*w) (where
the frequency vector w contains only positive frequencies).
To see the Bode plot of a transfer function, you can use the MATLAB
bode
command.

For example,

bode(50,[1 9 30 40])

displays the Bode plots for the

transfer function:

50 / (s^3 + 9 s^2 + 30 s + 40)

Gain and Phase Margin

Let's say that we have the following system:

where K is a variable (constant) gain and G(s) is the plant under consideration.

The gain margin is defined as the change in open loop gain required to make the
system unstable. Systems with greater gain margins can withstand greater changes
in system parameters before becoming unstable in closed loop. Keep in mind that
unity gain in magnitude is equal to a gain of zero in dB.

The phase margin is defined as the change in open loop phase shift required to
make a closed loop system unstable.

The phase margin is the difference in phase between the phase curve and -180 deg
at the point corresponding to the frequency that gives us a gain of 0dB (the gain
cross over frequency, Wgc).

Likewise, the gain margin is the difference between the magnitude curve and 0dB
at the point corresponding to the frequency that gives us a phase of -180 deg (the
phase cross over frequency, Wpc).
Gain and Phase Margin

-180
Gain and Phase Margin
We can find the gain and phase margins for a system directly, by using MATLAB.
Just enter the margin command.
This command returns the gain
and phase margins, the gain and
phase cross over frequencies, and
a graphical representation of these
on the Bode plot.

margin(50,[1 9 30 40])
Gain and Phase Margin

M agnitude:

db  G     20 log G  j   
Phase shift:

ps G     arg G  j     360  if  arg G  j     0  1  0 

180

Assume
K
K  2 G ( s) 

s ( 1  s)  1 
s

 3

Next, choose a frequency rangefor the plots (use powers of 10 for convenient plotting):

lowest frequency (in Hz):  start  .01 number of p oints: N  50

highest frequency (in Hz):  end  100

  start  1
step size: r  log 

 end  N
range for plot: i  0  N range variable: i r
 i   end 10 si  j  i
Gain and Phase Margin

Guess forcrossover frequency:  c  1

Solve for t he gain crossover frequency:

 
 c  root db G   c   c   c  1.193

Calculate thephase margin:

 
pm  ps G   c  180 pm  18.265 degrees

Gain Margin

Now using t he p hase angle plot, estimate the frequency at which the phase shift crosses 180 degrees

 gm  1.8

Solve for  at the phase shift point of 180 degrees:

  
 gm  root ps G   gm  180   gm 
 gm  1.732

Calculate thegain margin:


gm  db G   gm  gm  6.021
The Nyquist Stability Criterion
The Nyquist plot allows us also to predict the stability and performance of a closed-loop system by
observing its open-loop behavior. The Nyquist criterion can be used for design purposes regardless of open-
loop stability (Bode design methods assume that the system is stable in open loop). Therefore, we use this
criterion to determine closed-loop stability when the Bode plots display confusing information.

The Nyquist diagram is basically a plot of G(j* w) where G(s) is the open-loop transfer function and w is a
vector of frequencies which encloses the entire right-half plane. In drawing the Nyquist diagram, both
positive and negative frequencies (from zero to infinity) are taken into account. In the illustration below we
represent positive frequencies in red and negative frequencies in green. The frequency vector used in
plotting the Nyquist diagram usually looks like this (if you can imagine the plot stretching out to infinity):

However, if we have open-loop poles or zeros on the jw axis, G(s) will not be defined at those points, and
we must loop around them when we are plotting the contour. Such a contour would look as follows:
The Cauchy criterion

The Cauchy criterion (from complex analysis) states that when taking a closed contour in
the complex plane, and mapping it through a complex function G(s), the number of
times that the plot of G(s) encircles the origin is equal to the number of zeros of G(s)
enclosed by the frequency contour minus the number of poles of G(s) enclosed by the
frequency contour. Encirclements of the origin are counted as positive if they are in the
same direction as the original closed contour or negative if they are in the opposite
direction.

When studying feedback controls, we are not as interested in G(s) as in the closed-loop
transfer function:
G(s)
---------
1 + G(s)
If 1+ G(s) encircles the origin, then G(s) will enclose the point -1.
Since we are interested in the closed-loop stability, we want to know if there are any
closed-loop poles (zeros of 1 + G(s)) in the right-half plane.

Therefore, the behavior of the Nyquist diagram around the -1 point in the real axis is
very important; however, the axis on the standard nyquist diagram might make it hard
to see what's happening around this point
Gain and Phase Margin

Gain Margin is defined as the change in open-loop gain expressed in decibels (dB), required at 180
degrees of phase shift to make the system unstable. First of all, let's say that we have a system that
is stable if there are no Nyquist encirclements of -1, such as :
50
-----------------------
s^3 + 9 s^2 + 30 s + 40

Looking at the roots, we find that we have no open loop poles in the right half plane and therefore no
closed-loop poles in the right half plane if there are no Nyquist encirclements of -1. Now, how much
can we vary the gain before this system becomes unstable in closed loop?

The open-loop system represented by this plot will become unstable in closed loop if the gain is
increased past a certain boundary.
The Nyquist Stability Criterion

and that the Nyquist diagram can be viewed by typing:

nyquist (50, [1 9 30 40 ])
Gain and Phase Margin

Phase margin as the change in open-loop phase shift required at unity gain to make a
closed-loop system unstable.

From our previous example we know that this particular system will be unstable in closed
loop if the Nyquist diagram encircles the -1 point. However, we must also realize that if the
diagram is shifted by theta degrees, it will then touch the -1 point at the negative real axis,
making the system marginally stable in closed loop. Therefore, the angle required to make
this system marginally stable in closed loop is called the phase margin (measured in
degrees). In order to find the point we measure this angle from, we draw a circle with
radius of 1, find the point in the Nyquist diagram with a magnitude of 1 (gain of zero dB),
and measure the phase shift needed for this point to be at an angle of 180 deg.
 The Nyquist Stability Criterion

w  100 99.9 100 j  1 s ( w)  j w f ( w)  1

50 4.6
G( w) 
3 2
s ( w)  9 s ( w)  30 s ( w)  40
5

Im( G( w) )
0
0

5
2 1 0 1 2 3 4 5 6
Re( G( w) )
Consider the Negative Feedback System

Remember from the Cauchy criterion that the number N of times that the plot of G(s)H(s) encircles -1 is
equal to the number Z of zeros of 1 + G(s)H(s) enclosed by the frequency contour minus the number P
of poles of 1 + G(s)H(s) enclosed by the frequency contour (N = Z - P).
Keeping careful track of open- and closed-loop transfer functions, as well as numerators and
denominators, you should convince yourself that:
 the zeros of 1 + G(s)H(s) are the poles of the closed-loop transfer function
 the poles of 1 + G(s)H(s) are the poles of the open-loop transfer function.
The Nyquist criterion then states that:
 P = the number of open-loop (unstable) poles of G(s)H(s)
 N = the number of times the Nyquist diagram encircles -1
 clockwise encirclements of -1 count as positive encirclements
 counter-clockwise (or anti-clockwise) encirclements of -1 count as negative encirclements
 Z = the number of right half-plane (positive, real) poles of the closed-loop system
The important equation which relates these three quantities is:

Z = P + N
The Nyquist Stability Criterion - Application

Knowing the number of right-half plane (unstable) poles in open loop (P), and the
number of encirclements of -1 made by the Nyquist diagram (N), we can determine
the closed-loop stability of the system.

If Z = P + N is a positive, nonzero number, the closed-loop system is unstable.

We can also use the Nyquist diagram to find the range of gains for a closed-loop
unity feedback system to be stable. The system we will test looks like this:

where G(s) is :
s^2 + 10 s + 24
---------------
s^2 - 8 s + 15
 The Nyquist Stability Criterion

This system has a gain K which can be varied in order to modify the response of the closed-loop
system. However, we will see that we can only vary this gain within certain limits, since we have to
make sure that our closed-loop system will be stable. This is what we will be looking for: the range
of gains that will make this system stable in the closed loop.

The first thing we need to do is find the number of positive real poles in our open-loop transfer
function:

roots([1 -8 15])
ans =
5
3
The poles of the open-loop transfer function are both positive. Therefore, we need two anti-
clockwise (N = -2) encirclements of the Nyquist diagram in order to have a stable closed-loop
system (Z = P + N). If the number of encirclements is less than two or the encirclements are not
anti-clockwise, our system will be unstable.

Let's look at our Nyquist diagram for a gain of 1:

nyquist([ 1 10 24], [ 1 -8 15])

There are two anti-clockwise encirclements of -1.

Therefore, the system is stable for a gain of 1.
The Nyquist Stability Criterion

MathCAD Implementation

w  100 99.9 100 j  1 s ( w)  j w

2
s ( w)  10 s ( w)  24
G( w)  There are two anti-
2
s ( w)  8 s ( w)  15 clockwise encirclements of -
1.
2 Therefore, the system is
stable for a gain of 1.

Im( G( w) )
0
0

2
2 1 0 1 2
Re( G( w) )
The Nyquist Stability Criterion
Time-Domain Performance Criteria Specified
In The Frequency Domain

Open and closed-loop frequency resp onses are relat ed by :

G  j
T j
1  G  j
1
Mpw   0.707
2
2   1  

G   u  j v M M  

G  j
2 2
u  jv u v
M  
1  G  j 1  u  jv 2 2
( 1  u)  v

Squaring and rearrenging

2 which is t he equation of a
 M 
2
2  M 
2 circle on u-v planwe with a
u v 
 2 2 cent er at
 1M   1  M  M
2
u v 0
2
1M
Time-Domain Performance Criteria Specified
In The Frequency Domain
The Nichols Stability Method

Pol ar S tabi lity Plot - Nichol sMathcad Implementati on

This examp le makes a polar p lot of a transfer function and draws one contour of const ant
closed-loop magnitude. To draw the plot, enter a definition for the transfer function
G(s):

45000
G ( s) 
s ( s  2) ( s  30)
The frequency range defined by the next t wo equations provides a logarithmic frequency scale
and  m:
running from 1 to 100. You can change this range by editing the definitionsmfor

.02 m
m  0  100  m  10
Now enter a value forM to define the closed-loop magnitude contour t hat will be plot ted.
M  1.1
Calculate the points on the M -circle:

 M2 
M Cm  
M

exp 2  j .01 m
 M2  1 2 
 M 1 
The first plot shows
G, the contour of constant closed-loop magnit ude,
M
The Nichols Stability Method

The first plot shows

G, the contour of constant closed-loop magnitude,
M, and the
Nyquist of the open loop system

Im  G  j  m 

Im  MCm

Re  G  j  m   Re  MCm   1
The Nichols Stability Method
 Mpw
The Nichols Stability Method

G    
1
j   j   1  0.2  j   1

Mpw  2.5 dB  r  0.8

The closed-loop phase angle

at r is equal t o -72 degrees andb = 1.33
The closed-loop phase angle atb is equal to
-142 degrees

-3dB
-72 deg wr=0.8
-142 deg
The Nichols Stability Method

G    
0.64
j   j    j   1
2

Phase M argin = 30 degrees

  0.30
On the basis of the p hase we estimate

Mpw  9 dB Mpw  2.8  r  0.88 GM

From equation
1
Mpw   0.18
2
2   1  
s
We are confronted with comflecting

The app arent conflict is caused by the nature of

G(j) which slopes rap idally toward 180 degrees
line from the 0-dB axis. The designer must use PM
the frequency-domain-time-domain correlation
with caution
The Nichols Stability Method

GM

PM
Examples – Bode and Nyquist
Examples - Bode
Examples - Bode
Examples – Bode and Nyquist
Examples - Nichols
Examples - Nichols
The Design of Feedback Control Systems

PID

Compensation Networks
Different Types of Feedback Control

On-Off Control

This is the simplest form of control.

Proportional Control
A proportional controller attempts to perform better than the On-off type by
applying power in proportion to the difference in temperature between the
measured and the set-point. As the gain is increased the system responds faster
to changes in set-point but becomes progressively underdamped and eventually
unstable. The final temperature lies below the set-point for this system because
some difference is required to keep the heater supplying power.
Proportional, Derivative Control

The stability and overshoot problems that arise when a proportional

controller is used at high gain can be mitigated by adding a term
proportional to the time-derivative of the error signal. The value of the
damping can be adjusted to achieve a critically damped response.
Proportional+Integral+Derivative Control

Although PD control deals neatly with the overshoot and ringing

problems associated with proportional control it does not cure the
problem with the steady-state error. Fortunately it is possible to
eliminate this while using relatively low gain by adding an integral term
to the control function which becomes
The Characteristics of P, I, and D controllers

A proportional controller (Kp) will have the effect of reducing the rise time and will
reduce, but never eliminate, the steady-state error.

An integral control (Ki) will have the effect of eliminating the steady-state error, but it
may make the transient response worse.

A derivative control (Kd) will have the effect of increasing the stability of the system,
reducing the overshoot, and improving the transient response.
Proportional Control
By only employing proportional control, a steady state error occurs.

Proportional and Integral Control

The response becomes more oscillatory and needs longer to settle, the error
disappears.

Proportional, Integral and Derivative Control

All design specifications can be reached.
The Characteristics of P, I, and D controllers

CL RESPONSE RISE TIME OVERSHOOT SETTLING TIME S-S ERROR

Kp Decrease Increase Small Change Decrease

Ki Decrease Increase Increase Eliminate

Kd Small Change Decrease Decrease Small Change

Tips for Designing a PID Controller

1. Obtain an open-loop response and determine what needs to be

improved
2. Add a proportional control to improve the rise time
3. Add a derivative control to improve the overshoot
4. Add an integral control to eliminate the steady-state error
5. Adjust each of Kp, Ki, and Kd until you obtain a desired overall response.

Lastly, please keep in mind that you do not need to implement all three controllers
(proportional, derivative, and integral) into a single system, if not necessary. For
example, if a PI controller gives a good enough response (like the above
example), then you don't need to implement derivative controller to the system.
Keep the controller as simple as possible.
Open-Loop Control - Example

1
G( s )
2
s  10s  20

num=1;
den=[1 10 20];
step(num,den)
Proportional Control - Example

The proportional controller (Kp) reduces the rise time, increases the overshoot, and
reduces the steady-state error.

MATLAB Example
Kp
T( s )
2
s  10 s  ( 20  Kp)
Step Response
From: U(1)
1.4

Kp=300; 1.2 Step Response

From: U(1)
1

num=[Kp]; 1
0.9
Amplitude

0.8 0.8
den=[1 10 20+Kp];
To: Y(1)

0.7
0.6

0.6
t=0:0.01:2;
Amplitude

To: Y(1)
0.4
0.5

K=300 K=100
step(num,den,t) 0.2 0.4

0.3
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
0.2
Time (sec.)
0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (sec.)
 Proportional - Derivative - Example

The derivative controller (Kd) reduces both the overshoot and the settling time.

MATLAB Example

Kd s  Kp
T( s )
2
s  ( 10  Kd)  s  ( 20  Kp)
Step Response
From: U(1)
1.4

1.2

Kp=300; 1 Step Response

From: U(1)
1

Kd=10;
Amplitude

0.8
To: Y(1)

0.9

0.6 0.8

num=[Kd Kp]; 0.7

0.4
Kd=10 0.6

Amplitude
den=[1 10+Kd 20+Kp];
To: Y(1)
0.2 0.5

0.4
0

t=0:0.01:2; 0 0.2 0.4 0.6 0.8 1

Time (sec.)
1.2 1.4 1.6 1.8
0.3
2

0.2 Kd=20
step(num,den,t) 0.1

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Time (sec.)
Proportional - Integral - Example

The integral controller (Ki) decreases the rise time, increases both the overshoot and the
settling time, and eliminates the steady-state error

MATLAB Example
Kp s  Ki
T( s )
3 2
s  10 s  ( 20  Kp)  s  Ki
Step Response
From: U(1)
1.4

Kp=30;
1.2
Step Response
From: U(1)
1 1.4

Ki=70; 1.2
Amplitude

0.8
To: Y(1)

1
0.6

num=[Kp Ki];
Ki=70
Amplitude
0.8

To: Y(1)
0.4

den=[1 10 20+Kp Ki]; 0.2

0.6

0.4
0

t=0:0.01:2; 0 0.2 0.4 0.6 0.8 1

Time (sec.)
1.2 1.4 1.6 1.8
0.2
2
Ki=100

step(num,den,t) 0
0 0.2 0.4 0.6 0.8 1

Time (sec.)
1.2 1.4 1.6 1.8 2
RLTOOL

Syntax

rltool
rltool(sys)
rltool(sys,comp)
RLTOOL
RLTOOL
RLTOOL
RLTOOL
Example - Practice

Consider the following configuration:

Example - Practice

The design a system for the following specifications:

· Zero steady state error

· Settling time within 5 seconds
· Rise time within 2 seconds
· Only some overshoot permitted
Lead or Phase-Lead Compensator Using Root Locus

A first-order lead compensator can be designed using the root locus. A lead compensator
in root locus form is given by
( s  z)
G c( s )
( s  p)
where the magnitude of z is less than the magnitude of p. A phase-lead compensator
tends to shift the root locus toward the left half plane. This results in an improvement in
the system's stability and an increase in the response speed.

When a lead compensator is added to a system, the value of this intersection will be a
larger negative number than it was before. The net number of zeros and poles will be the
same (one zero and one pole are added), but the added pole is a larger negative number
than the added zero. Thus, the result of a lead compensator is that the asymptotes'
intersection is moved further into the left half plane, and the entire root locus will be
shifted to the left. This can increase the region of stability as well as the response speed.
Lead or Phase-Lead Compensator Using Root Locus

In Matlab a phase lead compensator in root locus form is implemented by using the
transfer function in the form
numlead=kc*[1 z];
denlead=[1 p];

and using the conv() function to implement it with the numerator and
denominator of the plant
newnum=conv(num,numlead);
newden=conv(den,denlead);
Lead or Phase-Lead Compensator Using Frequency Response

A first-order phase-lead compensator can be designed using the frequency response. A

lead compensator in frequency response form is given by

 1     s  1
Gc( s )
   1   s 
p
1

z
1

m z p  
sin m
1

In frequency response design, the phase-lead compensator adds positive phase to the
system over the frequency range. A bode plot of a phase-lead compensator looks like the
following
Lead or Phase-Lead Compensator Using Frequency Response

Additional positive phase increases the phase margin and thus increases the stability of
the system. This type of compensator is designed by determining alfa from the amount
of phase needed to satisfy the phase margin requirements, and determining tal to place
the added phase at the new gain-crossover frequency.

Another effect of the lead compensator can be seen in the magnitude plot. The lead
compensator increases the gain of the system at high frequencies (the amount of this
gain is equal to alfa. This can increase the crossover frequency, which will help to
decrease the rise time and settling time of the system.
Lead or Phase-Lead Compensator Using Frequency Response

In Matlab, a phase lead compensator in frequency response form is

implemented by using the transfer function in the form
numlead=[aT 1];
denlead=[T 1];
and using the conv() function to multiply it by the numerator and
denominator of the plant
newnum=conv(num,numlead);
newden=conv(den,denlead);
Lag or Phase-Lag Compensator Using Root Locus

A first-order lag compensator can be designed using the root locus. A lag compensator in root
locus form is given by
( s  z)
G c( s )
( s  p)
where the magnitude of z is greater than the magnitude of p. A phase-lag compensator tends
to shift the root locus to the right, which is undesirable. For this reason, the pole and zero of a
lag compensator must be placed close together (usually near the origin) so they do not
appreciably change the transient response or stability characteristics of the system.

When a lag compensator is added to a system, the value of this intersection will be a smaller
negative number than it was before. The net number of zeros and poles will be the same (one
zero and one pole are added), but the added pole is a smaller negative number than the
added zero. Thus, the result of a lag compensator is that the asymptotes' intersection is
moved closer to the right half plane, and the entire root locus will be shifted to the right.
Lag or Phase-Lag Compensator Using Root Locus

It was previously stated that that lag controller should only minimally change the
transient response because of its negative effect. If the phase-lag compensator is
not supposed to change the transient response noticeably, what is it good for?
The answer is that a phase-lag compensator can improve the system's steady-state
response. It works in the following manner. At high frequencies, the lag controller
will have unity gain. At low frequencies, the gain will be z0/p0 which is greater
than 1. This factor z/p will multiply the position, velocity, or acceleration constant
(Kp, Kv, or Ka), and the steady-state error will thus decrease by the factor z0/p0.
In Matlab, a phase lead compensator in root locus form is implemented by using
the transfer function in the form
numlag=[1 z];
denlag=[1 p];
and using the conv() function to implement it with the numerator and
denominator of the plant
newnum=conv(num,numlag);
newden=conv(den,denlag);
Lag or Phase-Lag Compensator using Frequency Response

A first-order phase-lag compensator can be designed using the frequency response. A

 1     s 
lag compensator in frequency response form is given by
G c( s )
   1   s 
The phase-lag compensator looks similar to a phase-lead compensator, except that a is
now less than 1. The main difference is that the lag compensator adds negative phase
to the system over the specified frequency range, while a lead compensator adds
positive phase over the specified frequency. A bode plot of a phase-lag compensator
looks like the following
Lag or Phase-Lag Compensator using Frequency Response

In Matlab, a phase-lag compensator in frequency response form is

implemented by using the transfer function in the form
numlead=[a*T 1];
denlead=a*[T 1];
and using the conv() function to implement it with the numerator and
denominator of the plant
newnum=conv(num,numlead);
newden=conv(den,denlead);
Lead-lag Compensator using either Root Locus or Frequency Response

A lead-lag compensator combines the effects of a lead compensator with those of a lag
compensator. The result is a system with improved transient response, stability and
steady-state error. To implement a lead-lag compensator, first design the lead
compensator to achieve the desired transient response and stability, and then add on a
lag compensator to improve the steady-state response
Exercise - Dominant Pole-Zero Approximations and Compensations

The influence of a particular pole (or pair of complex poles) on the response is mainly determined
by two factors: the real part of the pole and the relative magnitude of the residue at the pole. The
real part determines the rate at which the transient term due to the pole decays; the larger the real
part, the faster the decay. The relative magnitude of the residue determines the percentage of the
total response due to a particular pole.

Investigate (using S imulink) the impact of a closed-loop negative real pole on the overshoot of a
system having complex poles.

2
pr n
T( s )
( s  pr)  s   2  n  s   n 
2 2

 (0.3,
Make pr to vary (2, 3, 5) times the real part of the complex pole for different values of 0.5,
0.7).
Investigate (using S imulink) the impact of a closed-loop negative real zero on the overshoot of a
system having complex poles.

( s  zr)
T( s )
s 2   2  n  s   n 2
 

 (0.3,
Make zr to vary (2, 3, 5) times the real part of the complex pole for different values of 0.5, 0.7).
Exercise - Lead and Lag Compensation

Investigate (using Matlab and Simulink) the effect of lead and lag compensations on the two
systems indicated below. Summarize your observations. Plot the root-locus, bode diagram
and output for a step input before and after the compensations.

Remember
lead compensation: z<p (place zero below the desired root location or to the left of the first two
real poles)
lag compensation: z>p (locate the pole and zero near the origin of the s-plane)

Lead Compensation (use z=1.33, p=20 and K =15).

Lag Compensation (use z=0.09 , and p=0.015, K=1/6 )

Summarize your findings

Problem 10.36

Determine a compensator so that the percent overshoot is less than 20% and Kv
(velocity constant) is greater than 8.
 Poles and Zeros and Transfer Functions

Transfer Function: A transfer function is defined as the ratio of the Laplace

transform of the output to the input with all initial
conditions equal to zero. Transfer functions are defined
only for linear time invariant systems.

Considerations: Transfer functions can usually be expressed as the ratio

of two polynomials in the complex variable, s.

Factorization: A transfer function can be factored into the following form.

K ( s  z )( s  z ) ... ( s  z )
G( s)  1 2 m

( s  p )( s  p ) ... ( s  p )
1 2 n

The roots of the numerator polynomial are called zeros.

The roots of the denominator polynomial are called poles.

wlg
 Poles, Zeros and the S-Plane
An Example: You are given the following transfer function. Show the
poles and zeros in the s-plane.

( s  8)( s  14)
G(s) 
s( s  4)( s  10)

j axis

S - plane

origin
o x o x x
-14 -10 -8 -4 0  axis

wlg
 Poles, Zeros and Bode Plots

Characterization: Considering the transfer function of the

previous slide. We note that we have 4 different
types of terms in the previous general form:
These are:
1 1
K , , , ( s / z  1)
s ( s / p  1)
B

Expressing in dB: Given the tranfer function:

K ( jw / z 1)
G ( jw)  B

( jw)( jw / p 1)

20 log | G( jw |  20 log K  20 log | ( jw / z  1) | 20 log | jw | 20 log | jw / p  1 |

B

wlg
 Poles, Zeros and Bode Plots

Mechanics: We have 4 distinct terms to consider:

20logKB

20log|(jw/z +1)|

-20log|jw|

-20log|(jw/p + 1)|

wlg
 1 1 1 1 1 1

This is a sheet of 5 cycle, semi-log paper.

This is the type of paper usually used for
preparing Bode plots.

dB Mag
Phase
(deg)

wlg

 (rad/sec)
 Poles, Zeros and Bode Plots

Mechanics: The gain term, 20logKB, is just so many

dB and this is a straight line on Bode paper,
independent of omega (radian frequency).

The term, - 20log|jw| = - 20logw, when plotted

on semi-log paper is a straight line sloping at
-20dB/decade. It has a magnitude of 0 at w = 1.

20
-20db/dec

-20

=1
wlg
 Poles, Zeros and Bode Plots

Mechanics: The term, - 20log|(jw/p + 1), is drawn with the

following approximation: If w < p we use the
approximation that –20log|(jw/p + 1 )| = 0 dB,
a flat line on the Bode. If w > p we use the
approximation of –20log(w/p), which slopes at
-20dB/dec starting at w = p. Illustrated below.
It is easy to show that the plot has an error of
-3dB at w = p and – 1 dB at w = p/2 and w = 2p.
One can easily make these corrections if it is
appropriate.

2
0
0
-20db/dec
-20

-40

 =p wlg
 Poles, Zeros and Bode Plots
Mechanics: When we have a term of 20log|(jw/z + 1)| we
approximate it be a straight line of slop 0 dB/dec
when w < z. We approximate it as 20log(w/z)
when w > z, which is a straight line on Bode paper
with a slope of + 20dB/dec. Illustrated below.

20
+20db/dec
0

-20

-40

 =z

wlg
Example 1:

Given: 50, 000( jw  10)

G( jw) 
( jw  1)( jw  500)
First: Always, always, always get the poles and zeros in a form such that
the constants are associated with the jw terms. In the above example
we do this by factoring out the 10 in the numerator and the 500 in the
denominator.

50, 000 x10( jw /10  1) 100( jw /10  1)

G( jw)  
500( jw  1)( jw / 500  1) ( jw  1)( jw / 500  1)

Second: When you have neither poles nor zeros at 0, start the Bode
at 20log10K = 20log10100 = 40 dB in this case.

wlg
Example 1: (continued)

Third: Observe the order in which the poles and zeros occur.
This is the secret of being able to quickly sketch the Bode.
In this example we first have a pole occurring at 1 which
causes the Bode to break at 1 and slope – 20 dB/dec.
Next, we see a zero occurs at 10 and this causes a
slope of +20 dB/dec which cancels out the – 20 dB/dec,
resulting in a flat line ( 0 db/dec). Finally, we have a
pole that occurs at w = 500 which causes the Bode
to slope down at – 20 dB/dec.

We are now ready to draw the Bode.

Before we draw the Bode we should observe the range

over which the transfer function has active poles and zeros.
This determines the scale we pick for the w (rad/sec)
at the bottom of the Bode.

The dB scale depends on the magnitude of the plot and

experience is the best teacher here. wlg
Bode Plot Magnitude for 100(1 + jw/10)/(1 + jw/1)(1 + jw/500)

1 1 1 1 1 1
1 1 1 1 1 1
60

40

20

dB Mag Phase (deg)

dB Mag Phase (deg)
0

-20

-60

-60
0.1 1 10 100  (rad/sec)
1000 10000
 (rad/sec)

wlg
 Using Matlab For Frequency Response

Instruction: We can use Matlab to run the frequency response for

the previous example. We place the transfer function
in the form:
5000( s 10) [ 5000s  50000 ]

( s 1)( s  500) [ s  501s  500]
2

The Matlab Program

num = [5000 50000];
den = [1 501 500];
Bode (num,den)
In the following slide, the resulting magnitude and phase plots (exact)
are shown in light color (blue). The approximate plot for the magnitude
(Bode) is shown in heavy lines (red). We see the 3 dB errors at the
corner frequencies.

wlg
 Bode Diagrams

From: U(1)
40

30

20
Phase (deg); Magnitude (dB)

10

-10
1 10 100 500
0

-20
To: Y(1)

-40

-60
100(1  jw / 10)
Bode for: G ( jw) 
-80 (1  jw)(1  jw / 500)
-100
-1 0 1 2 3 4
10 10 10 10 10 10

Frequency (rad/sec)
wlg
 Phase for Bode Plots
Comment: Generally, the phase for a Bode plot is not as easy to draw
or approximate as the magnitude. In this course we will use
an analytical method for determining the phase if we want to
make a sketch of the phase.

Illustration: Consider the transfer function of the previous example.

We express the angle as follows:

G( jw)  tan (w / 10)  tan ( w / 1)  tan ( w / 500)

1 1 1

We are essentially taking the angle of each pole and zero.

Each of these are expressed as the tan-1(j part/real part)

Usually, about 10 to 15 calculations are sufficient to determine

a good idea of what is happening to the phase.

wlg
 Bode Plots
Example 2: Given the transfer function. Plot the Bode magnitude.

100(1  s / 10)
G( s) 
s(1  s / 100)
2

Consider first only the two terms of

100
jw
Which, when expressed in dB, are; 20log100 – 20 logw.
This is plotted below.

The is
40
-20db/dec a tentative line we use
dB 20 until we encounter the
first pole(s) or zero(s)
0
not at the origin.
-20

wlg
1  (rad/sec)
 Bode Plots
Example 2: (continued) The completed plot is shown below.
100(1  s / 10)
G( s) 
s(1  s / 100)2

1 1 1 1 1 1
60

-20db/dec
40

20

dB Mag -40 db/dec Phase (deg)

0

-20

100(1  s / 10)
G( s) 
-40 s(1  s / 100)
2

-60
0.1 1 10 100 1000
wlg
 (rad/sec)
 Bode Plots
Example 3:

80(1  jw)3
Given: G( s)  20log80 = 38 dB
( jw)3 (1  jw / 20) 2
1 1 1 1 1 1

-60 dB/dec

60

dB Mag 40

-40 dB/dec
20

-20 .

0.1 1 10 100 wlg

 (rad/sec)
 Bode Plots
Example 4:

10(1  jw / 2)
Given: G( jw) 
(1  j0.025w)(1  jw / 500) 2
1 1 1 1 1 1
60

40

+ 20 dB/dec
20 -40 dB/dec

dB Mag Phase (deg)

0

-20 Sort of a low

pass filter
-40

-60
2
0.1 1 10 100 1000 wlg

 (rad/sec)
 Bode Plots
Given:
(1  jw / 30) (1 jw / 100)
2 2

G ( jw) 
(1 jw / 2) (1 jw / 1700)
2 2

Example 5
1 1 1 1 1 1
60

40

20

dB Mag Phase (deg)

0

-40 dB/dec
-20 Sort of a low
pass filter
+ 40 dB/dec
-40

-60
0.1 1 10 100 1000 wlg

 (rad/sec)
 Bode Plots

Given: problem 11.15 text

Example 6
640( jw  1)(0.01 jw  1) 64( jw  1)(0.01 jw  1)
H ( jw)  
( jw) ( jw  10)
2
( jw) (0.1 jw  1) 2

.
-40dB/dec

40 .
-20db/dec

20 .
-40dB/dec
dB mag 0

-20 .
-20dB/dec

-40 .

0.01 0.1 1 10 100 1000 wlg

 Bode Plots

Design Problem: Design a G(s) that has the following Bode plot.

Example 7

40
30 dB

20
+40 dB/dec
-40dB/dec

0
dB mag
? ?

30 900
0.1 1 10 100 1000 wlg
 rad/sec
 Bode Plots
Procedure: The two break frequencies need to be found.
Recall:
#dec = log10[w2/w1]
Then we have:

(#dec)( 40dB/dec) = 30 dB

log10[w1/30] = 0.75 w1 = 5.33 rad/sec

Also:

log10[w2/900] (-40dB/dec) = - 30dB

This gives w2 = 5060 rad/sec

wlg
 Bode Plots
Procedure:

(1  s / 5.3)2 (1  s / 5060) 2
G( s) 
(1  s / 30)2 (1  s / 900)2
Clearing:
( s  5.3)2 ( s  5060)2
G( s) 
( s  30)2 ( s  900)2

Use Matlab and conv:

N1  ( s 2  10.6s  28.1) N 2  ( s 2  10120s  2.56 xe7 )

N1 = [1 10.6 28.1] N2 = [1 10120 2.56e+7]
N = conv(N1,N2)
1 1.86e+3 2.58e+7 2.73e+8 7.222e+8

s4 s3 s2 s1 s0

wlg
 Bode Plots
Procedure: The final G(s) is given by;

( s  10130.6s  2.571e s  2.716e s  7.194e )

4 3 8 2 8 8

G( s) 
( s  1860s  9.189e s  5.022e s  7.29e )
4 3 2 2 7 8

Testing: We now want to test the filter. We will check it at  = 5.3 rad/sec
And  = 164. At  = 5.3 the filter has a gain of 6 dB or about 2.
At  = 164 the filter has a gain of 30 dB or about 31.6.

We will check this out using MATLAB and particularly, Simulink.

wlg
Matlab (Simulink) Model:

wlg
Filter Output at  = 5.3 rad/sec

Produced from Matlab Simulink

wlg
Filter Output at  = 70 rad/sec

Produced from Matlab Simulink

wlg
 Reverse Bode Plot
Required:
From the partial Bode diagram, determine the transfer function
Example 8 (Assume a minimum phase system)

Not to scale
68

20 db/dec

-20 db/dec
30

20
db/dec
dB

1 110 850
 wlg
 Reverse Bode Plot
Required:
From the partial Bode diagram, determine the transfer function
Example 9 (Assume a minimum phase system)

100 dB
-40 dB/dec

Not to scale
50 dB
-20 dB/dec
-20 dB/dec

10 dB
-40 dB/dec

0.5 40 300
wlg

w (rad/sec)
Polar Plot
 Introduction

The polar plot of sinusoidal transfer function G(jω) is a plot of the

magnitude of G(jω) verses the phase angle of G(jω) on polar coordinates
as ω is varied from zero to infinity.
Therefore it is the locus of as ω is varied from zero to infinity.
As
So it is the plot of vector G jis) 
as( ω G( jfrom
varied ) zero to infinity

G( j) G( j)  Me j ( )

Me j ( )
 Introduction conti…

In the polar plot the magnitude of G(jω) is plotted as the distance from the
origin while phase angle is measured from positive real axis.
+ angle is taken for anticlockwise direction.
Polar plot is also known as Nyquist Plot.
Steps to draw Polar Plot

Step 1: Determine the T.F G(s)

Step 2: Put s=jω in the G(s)
Step 3: At ω=0 & ω=∞ find by &
Step 4: At ω=0 & ω=∞ find by &
G( j )
G( jthe
Step 5: Rationalize the function G(jω) and separate ) real andlim G( j )
imaginary
 0
partslim
 
Step 6: Put Re [G(jω) ]=0, determine the frequency at which plot intersects the Im axis and
calculate intersection value by putting the above ( j ) frequency
Gcalculated lim
 0
G( j )
in G(jω) lim
 
G( j )
Steps to draw Polar Plot conti…

Step 7: Put Im [G(jω) ]=0, determine the frequency at which plot intersects the
real axis and calculate intersection value by putting the above calculated
frequency in G(jω)
Step 8: Sketch the Polar Plot with the help of above information
Polar Plot for Type 0 System

K
Let G(s) 
(1  sT1 )(1  sT2 )

Step 1: Put s=jω

K
G ( j  )
Step 2: Taking the
(1  limit
jT1 )(for
1  magnitude
jT2 ) of G(jω)
K
   tan 1 T1  tan 1 T2
1  T1  1  T2 
2 2
Type 0 system conti…

K
lim G ( j )  K
1  T1  1  j T2 
2 2
0

K
Steplim G ( jthe
3: Taking ) limit of the Phase Angle of G(jω)
0
1  T1  1  j T2 
2 2
 

lim G ( j )    tan 1
T1  tan 1
T2  0
 0

lim G ( j )    tan 1
T1  tan 1 T2  180

Type 0 system conti…

Step 4: Separate the real and Im part of G(jω)

K (1   2T1T2 ) K (T1  T2 )
Step 5:G ( j ) 
Put Re [G(jω)]=0 j
1   T1   T2   T1T2
2 2 2 2 4
1   2T12   2T22   4T1T2

K (1   2T1T2 ) 1
 0    &
1   T1   T2   T1T2
2 2 2 2 4
T1T2
So When
1 K T1T2
  G ( j )    90 0
T1T2 T1  T2
&   G ( j )  0  180 0
Type 0 system conti…

Step 6: Put Im [G(jω)]=0

K (T1  T2 )
 0  0 & 
1   T1   T2   T1T2
2 2 2 2 4

So When
  0  G ( j )  K00
    G ( j )  01800
Type 0 system conti…
Polar Plot for Type 1 System

K
Let G(s) 
s(1  sT1 )(1  sT2 )

Step 1: Put s=jω

K
G ( j ) 
j (1  jT1 )(1  jT2 )
K
   90 0  tan 1 T1  tan 1 T2
 1  T1  1  j T2 
2 2
Type 1 system conti…

Step 2: Taking the limit for magnitude of G(jω)

K
lim G ( j  )  
 0  1  T1  1  j T2 
2 2

K
Step 3: Taking
lim G ( j  ) 
the limit of the Phase 0
Angle of G(jω)
   1  T1  1  j T2 
2 2

lim G ( j )    90 0
 tan 1
T1  tan 1
T2  90 0

 0

lim G ( j )    90 0
 tan 1 T1  tan 1 T2  270 0

Type 1 system conti…
Step 4: Separate the real and Im part of G(jω)

 K (T1  T2 ) j ( K 2 T1T2  K )
( j
Step 5: PutGRe )
[G(jω)]=0 j
   (T1  T2   T1 T2 )
3 2 2 2 2 2
   3 (T12  T22   2 T12 T22 )

 K (T1  T2 )
0 
   (T1  T2   T1 T2 )
3 2 2 2 2 2

So at
  G ( j )  0  270 0
Type 1 system conti…

Step 6: Put Im [G(jω)]=0

j ( K 2T1T2  K ) 1
 0    &   
   (T1  T2   T1 T2 )
3 2 2 2 2 2
T1T2
So When
1 K T1T2
  G ( j )   0 0
T1T2 T1  T2
  G ( j )  0 0
Type 1 system conti…
Polar Plot for Type 2 System
Let

Similar to above

K
G( s) 
s 2 (1  sT1 )(1  sT2 )
Type 2 system conti…
Note: Introduction of additional pole in denominator contributes a constant -
1800 to the angle of G(jω) for all frequencies. See the figure 1, 2 & 3
Figure 1+(-1800 Rotation)=figure 2
Figure 2+(-1800 Rotation)=figure 3
Ex: Sketch the polar plot for G(s)=20/s(s+1)(s+2)
Solution:
Step 1: Put s=jω

20
G ( j ) 
j ( j  1)( j  2)
20
   90 0  tan 1   tan 1  / 2
 2 1 2  4
Step 2: Taking the limit for magnitude of G(jω)

20
lim G ( j )  
0   1   4
2 2

20
lim G ( j )  0
   1   4
2 2

Step 3: Taking the limit of the Phase Angle of G(jω)

lim G ( j )    90 0
 tan 1
  tan 1
 / 2  90 0

 0

lim G ( j )    90 0
 tan 1   tan 1  / 2  270 0

Step 4: Separate the real and Im part of G(jω)

 60 2 j 20( 3  2 )
G( j )  4 j 4
(   )(4   )
2 2
(   2 )(4   2 )

 60 2
0 
(   )(4   )
4 2 2

So at
  G ( j )  0  270 0
Step 6: Put Im [G(jω)]=0

j 20( 3  2 )
 0     2 &   
(   )(4   )
4 2 2

So for positive value of 

10
  2 G ( j )   0 0
3
  G ( j )  00 0
Gain Margin, Phase Margin &
Stability
Phase Crossover Frequency (ωp) : The frequency where a polar plot intersects
the –ve real axis is called phase crossover frequency
Gain Crossover Frequency (ωg) : The frequency where a polar plot intersects
the unit circle is called gain crossover frequency
So at ωg

G( j )  Unity
Phase Margin (PM):
Phase margin is that amount of additional phase lag at the gain crossover frequency
required to bring the system to the verge of instability (marginally stabile)
Φm=1800+Φ
Where
Φ=∠G(jωg)
if Φm>0 => +PM (Stable System)
if Φm<0 => -PM (Unstable System)
 Gain Margin (GM):
The gain margin is the reciprocal of magnitudeG ( j )at the frequency at
which the phase angle is -1800.

In terms of dB

1 1
GM  
| G( jwc ) | x

1
GM in dB  20 log 10  20 log 10 | G( jwc ) | 20 log 10 ( x)
| G( jwc ) |
Stability

Stable: If critical point (-1+j0) is within the plot as shown, Both GM & PM are
+ve
GM=20log10(1
/x) dB
Unstable: If critical point (-1+j0) is outside the plot as shown, Both GM & PM
are -ve
GM=20log10(1
/x) dB
Marginally Stable System: If critical point (-1+j0) is on the plot as shown, Both
GM & PM are ZERO
GM=20log10(1
/1)=0 dB
MATLAB Margin
Inverse Polar Plot

The inverse polar plot of G(jω) is a graph of 1/G(jω) as a function of ω.

Ex: if G(jω) =1/jω then 1/G(jω)=jω

lim G ( j ) 1  0

0

lim G ( j ) 1  


Knowledge Before
Studying Nyquist Criterion

G( s)
T ( s) 
1  G( s) H ( s)

unstable if there is any pole on RHP (right half plane)

N G ( s) N H ( s)
G( s)  H ( s) 
DG ( s) DH ( s)
Open-loop system:
N G ( s) N H ( s)
G( s) H ( s) 
DG ( s) DH ( s)
Characteristic equation:

N G N H DG DH  N G N H
1  G( s) H ( s)  1  
DG DH DG DH

poles of G(s)H(s) and 1+G(s)H(s) are the same

Closed-loop system:

G( s) N G ( s) DH ( s)
T ( s)  
1  G( s) H ( s) DG ( s) DH ( s)  N G ( s) N H ( s)

zero of 1+G(s)H(s) is pole of T(s)

 ( s  1)(s  2)(s  3)(s  4)
G( s) H ( s) 
( s  5)(s  6)(s  7)(s  8)

G( s)
G( s) H ( s) 1  G( s ) H ( s )
1  G( s) H ( s)

Zero – 1,2,3,4 Zero – a,b,c,d Zero – ?,?,?,?

Poles – 5,6,7,8 Poles – 5,6,7,8 Poles – a,b,c,d

To know stability, we have to know a,b,c,d

Stability from Nyquist plot

From a Nyquist plot, we can tell a number of closed-loop poles on

the right half plane.
If there is any closed-loop pole on the right half plane, the system goes
unstable.
If there is no closed-loop pole on the right half plane, the system is
stable.
Nyquist Criterion

Nyquist plot is a plot used to verify stability

of the system.

mapping contour

( s  z1 )(s  z2 )
function F ( s) 
( s  p1 )(s  p2 )
mapping all points (contour) from one plane to another
by function F(s).
 ( s  z1 )(s  z2 )
F ( s) 
( s  p1 )(s  p2 )
Pole/zero inside the
contour has 360 deg.
angular change.
Pole/zero outside contour
has 0 deg. angular change.
Move clockwise around
contour, zero inside yields
rotation in clockwise, pole
inside yields rotation in
counterclockwise
Characteristic equation F ( s )  1  G( s ) H ( s )

N = P-Z
N = # of counterclockwise direction about the origin
P = # of poles of characteristic equation inside contour
= # of poles of open-loop system
z = # of zeros of characteristic equation inside contour
= # of poles of closed-loop system

Z = P-N
Characteristic equation
Increase size of the contour to cover the right half plane

More convenient to consider the open-loop system (with known pole/zero)

 Nyquist diagram of G( s) H ( s)
‘Open-loop system’

Mapping from characteristic equ. to open-loop

system by shifting to the left one step

Z = P-N

Z = # of closed-loop poles inside the right half plane

P = # of open-loop poles inside the right half plane
N = # of counterclockwise revolutions around -1
Properties of Nyquist plot

If there is a gain, K, in front of open-loop transfer function, the Nyquist plot

will expand by a factor of K.
Nyquist plot example

Open loop system has pole at 2

1
G( s) 
s pole
Closed-loop system has 2 at 1

If we multiply the open-loop with a

G( scan
gain, K, then we ) move 1 the closed-

 G( S ) to( sthe
loop pole’s 1position  1) left-half
plane
Nyquist plot example (cont.)

New look of open-loop system:

K
G( s) 
s2
Corresponding closed-loop system:

G( s) K

Evaluate value of 1 stability
K for G( s) s  ( K  2)

K 2
Adjusting an open-loop gain to guarantee stability

Step I: sketch a Nyquist Diagram

Step II: find a range of K that makes the system stable!
How to make a Nyquist plot?

Easy way by Matlab

Nyquist: ‘nyquist’
Bode: ‘bode’
Step I: make a Nyquist plot

Starts from an open-loop transfer function (set K=1)

Set and find frequency response
At dc,s  j
Find at which the imaginary part equals zero
 0s 0

 ( s  3)( s  5) s 2  8s  15
G ( s) H ( s)   2
( s  2)( s  4) s  6s  8
  2  8 j  15 (15   2 )  8 j
G ( j ) H ( j )  
   6 j  8
2
(8   2 )  6 j
(15   2 )  8 j (8   2 )  6 j
 
(8   )  6 j (8   2 )  6 j
2

(15   2 )(8   2 )  48 2  j (154  14 3 )


(8   2 ) 2  6 2  2
Need the imaginary term = 0,   0, 11

Substitute back
 in11
to the transfer function
And get
G(s)  1.33

(15  11)(8  11)  48(11)  540

  1.31
(8  11)  6 (11)
2 2
412
 At dc, s=0,

At imaginary part=0
Step II: satisfying stability condition

P = 2, N has to be 2 to guarantee stability

Marginally stable if the plot intersects -1
For stability, 1.33K has to be greater than 1

K > 1/1.33

or K > 0.75
Example
Evaluate a range of K that makes the system stable

K
G( s) 
( s 2  2s  2)(s  2)
Step I: find frequency at which imaginary part = 0

Set s  j
K
G ( j ) 
(( j ) 2  2 j  2)( j  2)
4(1   2 )  j (6   2 )

16(1   2 ) 2   2 (6   2 ) 2

At   0, 6 the imaginary part = 0

Plug  back6 in the transfer function

and get G = -0.05
Step II: consider stability condition

P = 0, N has to be 0 to guarantee stability

Marginally stable if the plot intersects -1
For stability, 0.05K has to be less than 1

K < 1/0.05

or K < 20
Gain Margin and Phase Margin

Gain margin is the change in open-loop gain (in dB),

required at 180 of phase shift to make the closed-loop
system unstable.

Phase margin is the change in open-loop phase shift,

required at unity gain to make the closed-loop
system unstable.

GM/PM tells how much system can tolerate

before going unstable!!!
GM and PM via Nyquist plot
GM and PM via Bode Plot

•The frequency at which the

phase equals 180 degrees is
called the phase crossover
frequency G
M

 GM
•The frequency at which the
magnitude equals 1 is called
the gain crossover frequency
 M

gain crossover frequency phase crossover frequency

Example

Find Bode Plot and evaluate a value of K

that makes the system stable
The system has a unity feedback
with an open-loop transfer function

K
G( s) 
( s  2)( s  4)(s  5)

First, let’s find Bode Plot of G(s) by assuming

that K=40 (the value at which magnitude plot
starts from 0 dB)
At phase = -180, ω = 7 rad/sec, magnitude = -20 dB
GM>0, system is stable!!!
Can increase gain up 20 dB without causing instability (20dB = 10)
Start from K = 40
with K < 400, system is stable
Closed-loop transient and closed-loop frequency
responses
‘2nd system’

C ( s) n2
 T ( s)  2
R( s ) s  2 n s  n2
Damping ratio and closed-loop frequency response

1
Mp 
2 1   2

 p  n 1  2 2

Magnitude Plot of closed-loop system

Response speed and closed-loop frequency response

 BW   n (1  2 2 )  4 4  4 2  2
4
 BW  (1  2 2 )  4 4  4 2  2
Ts

 BW  (1  2 2 )  4 4  4 2  2
Tp 1   2

 BW= frequency at which magnitude is 3dB down

from value at dc (0 rad/sec), or . 1
M
2
Find from  BW
Open-loop Frequency Response

Nichols Charts

From open-loop frequency response, we can find

 BW
at the open-loop frequency that the magnitude
lies between -6dB to -7.5dB (phase between -135 to -225)
Relationship between
damping ratio and phase margin
of open-loop frequency response
Phase margin of open-loop frequency response
Can be written in terms of damping ratio as following

2
M  tan 1

 2 2  1  4 4
Open-loop system with a unity feedback has a bode plot
below, approximate settling time and peak time

 BW= 3.7

PM=35
 2
M  tan 1

 2 2  1  4 4
Solve for PM = 35   0.32

4
Ts  (1  2 2 )  4 4  4 2  2
 BW 
 5.5

Tp  (1  2 2 )  4 4  4 2  2
 BW 1   2
 1.43
 UNIT-5

STATE SPACE REPRESENTATION

Objectives

How to find mathematical model, called a state-space representation, for a

linear, time-invariant system
How to convert between transfer function and state space models
How to find the solution of state equations for homogeneous &non
homogeneous systems
 Plant

Mathematical Model :
Linear, time invariant
Differential equation

Frequency Domain Time Domain

Technique Technique
Two approaches for analysis and design of control system

1. Classical Technique or Frequency Domain Technique

2. Modern Technique or Time Domain Technique
Some definitions

•System variable : any variable that responds to an input or initial conditions in a

system
•State variables : the smallest set of linearly independent system variables such that
the values of the members of the set at time t0 along with known forcing functions
completely determine the value of all system variables for all t ≥ t0
•State vector : a vector whose elements are the state variables
•State space : the n-dimensional space whose axes are the state variables
•State equations : a set of first-order differential equations with b variables, where
the n variables to be solved are the state variables
•Output equation : the algebraic equation that expresses the output variables of a
system as linear combination of the state variables and the inputs.
•For nth-order, write n simultaneous, first-order differential equations in terms of
the state variables (state equations).
•If we know the initial condition of all of the state variables at as well as the
system input for , we can solve the equations
Graphic
representation of
state space and a
state vector
 For a dynamic system, the state of a system is described in terms of a set of state
variables

[x1 (t ) x 2 (t )  x n (t )]
The state variables are those variables that determine the future behavior of a
system when the present state of the system and the excitation signals are known.
Consider the system shown in Figure 1, where y1(t) and y2(t) are the output signals
and u1(t) and u2(t) are the input signals. A set of state variables [x1 x2 ... xn] for the
system shown in the figure is a set such that knowledge of the initial values of the
state variables [x1(t0) x2(t0) ... xn(t0)] at the initial time t0, and of the input signals
u1(t) and u2(t) for t˃=t0, suffices to determine the future values of the outputs and
state variables.

Input Signals x(0) Initial conditions

Output Signals
u1(t) y1(t)
u(t) y(t)
u2(t) System y2(t) System
Input Output
Figure 1. Dynamic system.
In an actual system, there are several choices of a set of state variables that specify
the energy stored in a system and therefore adequately describe the dynamics of
the system.

The state variables of a system characterize the dynamic behavior of a system. The
engineer’s interest is primarily in physical, where the variables are voltages,
currents, velocities, positions, pressures, temperatures, and similar physical
variables.

The State Differential Equation:

The state of a system is described by the set of first-order differential equations
written in terms of the state variables [x1 x2 ... xn]. These first-order differential
equations can be written in general form as

x 1  a11x1  a12x 2   a1n x n  b11u1   b1m u m

x 2  a 21x1  a 22x 2   a 2 n x n  b 21u1   b 2 m u m

x n  a n1x1  a n 2 x 2   a nn x n  b n1u1   b nmu m
Thus, this set of simultaneous differential equations can be written in matrix form as
follows:

 x1   a11 a12  a1n   x1 

       b11  b1m   u1 
 a 2n  x 2  
        
d  x 2  a 21 a 22

dt           
       b n1  b nm  u m 
 x n  a n1 a n2  a nn   x n 

n: number of state variables, m: number of inputs.

The column matrix consisting of the state variables is called the state vector and is
written as

 x1 
x 
x   2
 
 
x n 
The vector of input signals is defined as u. Then the system can be represented by the
compact notation of the state differential equation as

x  A x  B u
This differential equation is also commonly called the state equation. The matrix A is
an nxn square matrix, and B is an nxm matrix. The state differential equation relates
the rate of change of the state of the system to the state of the system and the input
signals. In general, the outputs of a linear system can be related to the state variables
and the input signals by the output equation

y C xDu
Where y is the set of output signals expressed in column vector form. The state-space
representation (or state-variable representation) is comprised of the state variable
differential equation and the output equation.
General State Representation

x  Ax  Bu State equation

y  Cx  Du output equation

x = state vector

x = derivative of the state vector with respect to time

y = output vector
u = input or control vector
A = system matrix
B = input matrix
C = output matrix

D = feedforward matrix
AN EXAMPLE OF THE STATE VARİABLE CHARACTERİZATİON OF A SYSTEM

iL
ic L
u(t) R Vo
Vc
C
Current
source

• The state of the system can be described in terms of a set of variables [x1 x2],
where x1 is the capacitor voltage vc(t) and x2 is equal to the inductor current iL(t).
This choice of state variables is intuitively satisfactory because the stored energy
of the network can be described in terms of these variables.
Therefore x1(t0) and x2(t0) represent the total initial energy of the network and thus the
state of the system at t=t0.

Utilizing Kirchhoff’s current low at the junction, we obtain a first order differential
equation by describing the rate of change of capacitor voltage

dv c
ic  C  u(t)  i L
dt
Kirchhoff’s voltage low for the right-hand loop provides the equation describing the
rate of change of inducator current as

di L
L  R i L  v c
dt
The output of the system is represented by the linear algebraic equation

v0  R i L (t )
We can write the equations as a set of two first order differential equations in terms
of the state variables x1 [vC(t)] and x2 [iL(t)] as follows:

dv c dx1 1 1
C  u(t)  i L   x 2  u(t)
dt dt C C
di L dx 2 1 R
L  R i L  v c  x1  x 2
dt dt L L

The output signal is then y1 (t )  v0 ( t )  R x 2

Utilizing the first-order differential equations and the initial conditions of the
network represented by [x1(t0) x2(t0)], we can determine the system’s future and its
output.

The state variables that describe a system are not a unique set, and several
alternative sets of state variables can be chosen. For the RLC circuit, we might
choose the set of state variables as the two voltages, vC(t) and vL(t).
We can write the state variable differential equation for the RLC circuit as

 1
0   1
x   C x    u(t)
1 R C
0
    
L L
and the output as

y  0 R x
RLC network

1. State variables i t  qt 

2.  Ri   idt  vt 
di 1
L
dt C

Using it   dq dt
d 2q
L 2  R  q  vt 
dq 1
dt dt C

dq
i
dt
(1)
q  i  vt 
di 1 R 1

dt LC L L

3. qt  i t Can be solved using Laplace Transform

4. Other network variables can be obtained

vL t    qt   Ri t   vt 
1 (2)
C
5. (1),(2) : state-space representation
Other variables vR t  vC t 

  vR  vC  vt 
dvR R R R
dt L L L
dvC 1
 vR
dt RC

Each variables : linearly independent

In vector-matrix form

x  Ax  Bu (1)

where
dq dt   0 1 
x    A   1 / LC  R / L
 di dt   
q   0 
x  B  u  vt 
i  1 L 

y  Cx  Du (2)

where
y  vL t  C   1 C  R D 1
 State space representation using phase variable form

dny d n 1 y dy
n
 a n 1 n 1
   a1  a0 y  b0u
dt dt dt
choose
d n 1 y
x1  y x2 
dy
dt
d2y
x3  2
dt  xn  n 1
dt
diferensiasikan
dy
x1  x1  x2
dt
d2y
x2  2
x 2  x3
dt

d3y
x3  3 x n 1  xn
dt
 n
xn  a0 x1  a1 x2   an1 xn  b0u
d y
xn 
dt n
 x1   0 1 0 0 0 0  0   x1  0 
 x   0 0 1 0 0 0  0   x2  0 
 2  
 x3   0 0 0 1 0 0  0   x3  0 
      u
          
 xn 1   0 0 0 0 0 0  1   xn 1  0 
      
 xn   a0  a1  a2  a3  a4  a5   an 1   xn  b0 

 x1 
x 
 2 
 x3 
y  1 0 0  0 
  
 xn 1 
 
 xn 
Example : TF to State Space

1. Inverse Laplace
c  9c  26c  24c  24r

2. Select state variables

x1  c x2  c x3  c

x1  x2
x2  x3
N s 
G s  
x3  24 x1  26 x2  9 x3  24r Ds 

N s  numerator
y  c  x1 Ds  denominator
 x1   0 1 0   x1   0 
 x    0   x    0 r  x1 
y  1 0 0 x2 
  2 0 1  2   
 x3   24  26  9  x3  24
 x3 
Decomposing a transfer function
  
Y s   C s   b2 s 2  b1s  b0 X1 s 

d 2 x1
yt   b2 2  b1 1  b0 x1
dx
dt dt

yt   b0 x1  b1 x2  b2 x3
Example
yt   2 x1  7 x2  x3

 x1 
y  2 7 1 x2 
 x3 
State Space to TF

x  Ax  Bu
y  Cx  Du
Laplace Transform
sX s   AX s   BU s 
Y s   CX s   DU s 

X s   sI  A BU s 
1


Y s   C sI  A B  D U s 
1

Y s 
T s    C sI  A B  D
1

U s 
 0 1 0  10
x   0 0 1  x   0 u
 1  2  3  0 
y  1 0 0x
s 1 0 
sI  A  0 s  1 
1 2 s  3

sI  A1  adj (sI  A)

det(sI  A)

( s 2  3s  2) s3 1
 
  1 s ( s  3) s 
 s  (2s  1) s 
2

 
s 3  3s 2  2s  1
T s   C sI  A B  D
1

10( s 2  3s  2)
T s   3
s  3s 2  2s  1
Solution of homogeneous 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
  ax; x (0)  0
x
dx  axdt
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
x
on integrating above equation
logx=at+c
X= eat. ec
x=x(0)= ec
on substituting the intial condition ,the solution of homogeneous state equation of first order
differential equation is
x  e x(0)
at

x  AX (t ), x(o)  0
A2t 2 Ak t k
e At  I  At   
2! k!
Solution of homogeneous 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
  ax  bu; x(0)  0
x

Where x(t) and u(t) are scalar functions of time. By taking laplace transform

s X (s)  x0  a X (s)  bU (s)

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


a ( t )
x ( t )  e x (0) 
at
e 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

A2t 2 Ak t k
e At
 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 1

where we note that [sI-A]-1=ϕ(s), which is the Laplace transform of ϕ(t)=eAt. The
matrix exponential function ϕ(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
THE TRANSFER FUNCTION FROM THE STATE EQUATION

The transfer function of a single input-single output (SISO) system can be obtained
from the state variable equations.

x  A x  B u
yCx
where y is the single output and u is the single input. The Laplace transform of the
equations

sX(s)  AX(s)  B U(s)

Y(s)  CX(s)

where B is an nx1 matrix, since u is a single input. We do not include initial

conditions, since we seek the transfer function. Reordering the equation
 [sI  A] X(s)  B U(s)
X(s)  sI  A  BU(s)  (s)BU(s)
1

Y(s)  C(s)BU(s)
Therefore, the transfer function G(s)=Y(s)/U(s) is

G(s)  C(s)B
Example:
Determine the transfer function G(s)=Y(s)/U(s) for the RLC circuit as described by the
state differential function

 1
0   1
x   C x u y  0 R  x
R C ,

1
  0
 
L L
  R 1
1 s  L  
(s)  sI  A  
1 C
 1   
 s (s)  1 s 
sI  A   1 C 
R  L 
 s 
 L L R 1
(s)  s 2  s 
L LC

Then the transfer function is

 R 
 s 
L  1  1
 
G (s)  0 R  (s) C(s)   C 
 
 1 s  0
 L(s) (s) 
R / LC R / LC
G (s)  
(s) s  s
2 R 1
L LC
CONSIDER THE SYSTEM

C (s) 24
 3
R( s) s  9 s 2  26 s  24
s 3

 9 s 2  26 s  24 C ( s )  24 R ( s )
c  9c
  26c  24c  24r
State variables
x1  c
x2  c
x3  c  System
x1  x2
equations
x 2  x3
x3  - 24 x1 - 26 x2  9 x3  24r
y  c  x1
Output equation
x1  x2
x2  x3
x2  - 24 x1 - 26 x2  9 x3  24r
y  c  x1
 x1   0 1 0   x1   0 
 x    0 0 1   x    0 r
 2   2   
 x 2   24  26  9  x3  24
 x1 
y  1 0 0 x2 
 x3 
THANK YOU