Sem.

2
Control System Characteristics 2015/2016

Sensitivity, set-point tracking, and disturbance rejection

 Classical feedback control system

where : Unknown disturbance input, : Measurement noise

Therefore

 The system output is expressed as

 The sensitivity function that relates and disturbance input when

is given by

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 1

 Sem. 2
Sensitivity, tracking, and disturbance rejection 2015/2016

 Complementary sensitivity function is expressed as

 When , the output of the system is given by

 Therefore if and , a perfect set-point tracking and disturbance

rejection are realized
 The is required to be strictly perfect, so that

 If , perfect set-point tracking and noise acceptance are attained

 In frequency domain: at low frequencies good set-point tracking is
achieved, while at high frequencies good noise rejection is a realized

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 2

 Sem. 2
Stability Analysis 2015/2016

Stability (definition):
 BIBO (Bounded-Input-Bounded-Output) stability : Any bounded input
generates a bounded output.

 Asymptotic stability: For any initial condition, the output response converges
to zero.

For a system represented by a transfer function is stable if all the poles are in
the open left half of the complex plane
 Marginally stability: A dynamic system is said to be marginally stable if
 The transfer function has no poles in the Right Half Plane (RHP)

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 3

 Sem. 2
Stability Analysis Cont’ 2015/2016

 The transfer function has at least one simple pole on the imaginary axis of s-
plane
 The transfer function has no multiple poles of the imaginary axis of the s-plane
e.g.,

 Instability: A system is said to be unstable if it is neither stable nor marginally

stable
General remarks on the stability
 For a general system (nonlinear, etc), BIBO stability does not imply asymptotic
stability condition
 For a linear-time-invariant (LTI) system, BIBO stability implies asymptotic stability
condition
 For a LTI system, the term stability can be used to refer to both BIBO stability and
asymptotic stability
 Marginally stability does not imply BIBO stability
 Unstable system can be stabilized using feedback control
 Unstable closed-loop system is useless

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 4

 Sem. 2
Routh Hurwitz Stability Criterion 2015/2016

 Employed to determine stability of a dynamic system whose transfer function

contains no functions (sin, cos, exponential, etc.) in its polynomial denominator
 It determines the number of poles on the open Right Half Plane (RHP).
 It determines whether all the poles are on the Right Half Plane (LHP)
 Considering a characteristic equation , the Routh-
Hurwitz stability criterion states
 A necessary but not sufficient stability condition is that all the coefficients of the
characteristic equation must have the same sign and that none is zero
 All the Hurwitz determinant of the polynomial determinant must be positive
 All the coefficients of the first column of Routh’s array must have the same sign
 The Hurwitz determinants are given by

 Routh array can be written as shown in table below

where

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 5

 Sem. 2
Routh Hurwitz Stability Criterion Cont’ 2015/2016

 Routh array
The number of roots in tthe
open right half-plane is equal
to the number of sign changes
in the first column of Routh
array

 The array order can also be expressed in the reverse order

Example 4.1 (Burns)
Determine the number of poles with positive real part of the following characteristic
equations

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 6

 Sem. 2
Routh Hurwitz Stability Criterion Cont’ 2015/2016

Example 4.2
For the dynamic system described by the following closed-loop transfer function

a) Determine the characteristic polynomial

b) Generate the Routh table
c) Determine the stability
 Special cases of Routh array
 A zero in the first column: Replace a zero with a small arbitrary number and
proceed as usual
Example 4.3
Determine the stability of a closed-loop system described by the following transfer
function

 All elements in a row are zeros: The zeros are replaced by the derivatives of an
auxiliary polynomial formed from elements of the previous row, then proceed as
usual
Example 4.4
Determine the stability of a closed-loop system that have the following
characteristic equation

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 7

 Sem. 2
Routh Hurwitz Stability Criterion Cont’ 2015/2016

Example 4.4
Design a controller that stabilizes the following closed-loop system if is a
constant and when

Example 4.5

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 8

 Sem. 2
Root locus: Analysis and Design 2015/2016

System poles and zeros

 Consider a closed-loop transfer function

where are the closed-loop poles and are the

closed-loop zeros

 The position of closed-loop poles in s-plane determines the transient response

behavior of a dynamic system

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 9

 Sem. 2
Root locus: Analysis and Design Cont’ 2015/2016

Definition
Root locus is a plot of the roots of a system characteristic equation (or poles of
closed-loop transfer function) as the open-loop gain constant is varied from zero
to infinity
Example 4.6
Consider the following closed-loop system

Closed-loop transfer function

Characteristic equation

Poles

 The performance and the stability of the system are affected by changing the value
of .
Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 10
 Sem. 2
Root locus: Analysis and Design Cont’ 2015/2016

 The effects of varying on the natural frequency and damping ratio can easily be
determined

 Varying from zero infinity

Properties of root locus

 For a given closed-loop transfer function, the characteristic equation is given by

 Magnitude property: the magnitude of the open-loop transfer function is

expressed as

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 11

 Sem. 2
Root locus: Analysis and Design Cont’ 2015/2016

 Angle property: the angle of the open-loop transfer function is expressed as an

odd multiple of 180°, i.e.,

Sketching of the root locus

 The sketching of root locus (RL) can be realized by using angle and magnitude
criteria
 It is difficult to draw RL analytically for each and every value of
 For a precise sketching of RL, MATLAB program is used

For the steps to sketching RL, refer to the given handout

Example 4.7
Sketch the RL for the given open-loop transfer function (MATLAB)

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 12

 Sem. 2
Closed-loop control design using RL method 2015/2016

 Aims at placing the closed-loop poles at the desired location in order to achieve
given performance specifications by tuning the controller .
 If the root locus does not pass through the desired position, locus reshaping is
carried out by adding poles/zeros to the controller .
Effects of adding poles
 Pulls the root locus to the RIGHT
 Slows down the settling time
 Less stable

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 13

 Sem. 2
Closed-loop control design using RL method Cont’ 2015/2016

Effects of adding zeros

 Pulls the root locus to the LEFT
 Speeds up the settling time
 More stable

 Adding zeros alone leads to amplification of high frequency noise

 Adding poles alone may generate a less stable controller by moving the closed-loop
poles more to the right

Dr.-Ing. Jackson G. Njiri EMT 2339 L-4: Control System Characteristics 14