Stability of Control System

Introduction

The stability of a control system is defined as the ability of any system to

provide a bounded output when a bounded input is applied to it. Stability
allows the system to reach the steady-state and remain in that state for
that particular input even after variation in the parameters of the system.
Stability is considered to be an important property of a control system. It
is also referred as the system’s ability to reach the steady-state.
As it is an important characteristic thus the performance of the control
system shows a high dependency on stability.
Various units of the control system combinedly operate so as to provide a
specific output of the system.
In order to get the specified output, the various parameters of the system
must be controlled. Along with this, the system must be stable enough so
that the output must not get affected by the undesirable variations in the
parameter of the system or disturbances.

Thus we can say that a stable system is designed so as to get the desired
response of the system without any intolerable variation with the changes
in the system parameters.

It is to be noted here that stability or instability is the characteristic

property of the control system and thus depends on the closed-loop poles
of the system.

Therefore, we can say that stability is a factor of the system which is

independent of the input of the system. However, the steady-state output
of the system is dependent on the poles of the applied input.

Stability of Control System

A stable system generates bounded output for bounded input (BIBO).

Bounded value of a signal represents a finite value. The bounded signal
holds a finite value of maxima and minima. Thus, if maxima and minima
of any signal are finite then all the other values between maxima and
minima will also be finite.

Suppose we have a signal shown. This signal

has a maximum and minimum finite values.
This signal is said to be bounded and if such an
output is provided by a system then it is said to
be a stable system.

Therefore, conversely, we can say that an

unstable system provides an unbounded output
when the applied input is bounded in nature.
The signals whose value continuously rises, such as ramp signal and
raising exponential signals are known as unbounded signals.

Types of Stable System

 Absolutely stable system: An absolutely stable system is the one

that provides bounded output even for the variation in the
parameters of the system. This means it is such a system whose
output after reaching a steady-state does not show changes
irrespective of the disturbances or variation in the system parameter
values.
 Conditionally stable system: A conditionally stable system gives
bounded output for the only specific conditions of the system that is
defined by the parameter of the system. W e can say here the system
exhibits stability only under particular conditions. If that particular condition
is violated then the system generates unbounded output.

 Marginally/ Critically stable system: A marginally stable system

is the one that generates a signal which is oscillating with constant
frequency and amplitude when a bounded input is provided to it.
These oscillations are known as sustained oscillations. The figure here
represents the step response of a marginally stable system:

BIBO Stability

Consider a linear, time invariant system having an impulse

response h(t) and an input function of r(t)
The output of the system is given by,
t
c ( t ) =∫ h ( τ ) r ( t−τ ) dτ
0

For the system, h(t) = L [ T ( s ) ]=L

−1 −1 C (s)
R(s) [ ]
If the input is bounded, i.e., |r (t )|≤ R 1, we have

| |
t

|c(t )|= ∫ h ( τ ) r ( t−τ ) dτ

0
t
≤∫ |h ( τ ) r ( t−τ )|dτ
0
t
≤∫ |h ( τ )||r ( t−τ )| dτ
0
t
≤ R1∫|h ( τ )|dτ
0

For a stable system, bounded input should produce bounded

t

output. Thus the system is BIBO stable, if ∫|h ( τ )|dτ is finite or h(t)
0
 t

is absolutely integrable. If ∫|h ( τ )|dτ is plotted with respect to time,

0

this condition means that the area bounded by this curve and
time axis must be finite between the limits t = 0 and t = ∞ .
Thus the stability of the system can be ascertained from the
impulse response or the natural response of the system which
is independent of the input.
Since the nature of impulse response h(t) depends on the
location of poles of T(s), the transfer function T(s) can be
written as,
N (s)
T ( s )=
D (s)

D(s) = 0 is known as the characteristic equation of the system

and the poles of T(s) are the roots of D(s) = 0. The roots of D(s)
may be classified as:
i. Real roots at s = σ
ii. Complex conjugate roots at s = σ ± jω
iii. Roots at origin
iv. Purely imaginary roots at s = ± jω
The location of the roots and corresponding responses are
as shown in the figure:

.
Routh Stability Criterion
This criterion is also known as modified Hurwitz Criterion of stability of the
system. We will study this criterion in two parts. Part one will cover
necessary condition for stability of the system and part two will cover the
sufficient condition for the stability of the system. Let us again consider
the characteristic equation of the system as

1) Part one (necessary condition for stability of the system): In this we

have two conditions which are written below:
1. All the coefficients of the characteristic equation should be positive
and real.
2. All the coefficients of the characteristic equation should be non zero.
2) Part two (sufficient condition for stability of the system): Let us first
construct routh array. In order to construct the routh array follow these
steps:
 The first row will consist of all the even terms of the characteristic
equation. Arrange them from first (even term) to last (even term).
The first row is written below: a0 a2 a4 a6…………
 The second row will consist of all the odd terms of the characteristic
equation. Arrange them from first (odd term) to last (odd term). The
first row is written below: a1 a3 a5 a7………..
 The elements of third row can be calculated as:
(1) First element : Multiply a0 with the diagonally opposite element
of next column (i.e. a3) then subtract this from the product of a1 and
a2 (where a2 is diagonally opposite element of next column) and
then finally divide the result so obtain with a 1. Mathematically we
write as first element

(2) Second element: Multiply a0 with the diagonally opposite element of

next to next column (i.e. a5) then subtract this from the product of a1 and
a4 (where, a4 is diagonally opposite element of next to next column) and
then finally divide the result so obtain with a 1. Mathematically we write as
second element

Similarly, we can calculate all the elements of the third row.

(d) The elements of fourth row can be calculated by using the following
procedure:
(1) First element : Multiply b1 with the diagonally opposite element of next
column (i.e. a3) then subtract this from the product of a1 and b2 (where, b2
is diagonally opposite element of next column) and then finally divide the
result so obtain with b1. Mathematically we write as first element
(2) Second element : Multiply b1 with the diagonally opposite element of
next to next column (i.e. a5) then subtract this from the product of a1 and
b3 (where, b3 is diagonally opposite element of next to next column) and
then finally divide the result so obtain with a 1. Mathematically, we write as
second element

Similarly, we can calculate all the elements of the fourth row.

Similarly, we can calculate all the elements of all the rows.
Stability criteria if all the elements of the first column are positive then the
system will be stable. However if anyone of them is negative the system
will be unstable.
Now there are some special cases related to Routh Stability Criteria which
are discussed below:
1) Case one: If the first term in any row of the array is zero while the rest
of the row has at least one non zero term.
In this case we will assume a very small value (ε) which is tending to zero
in place of zero. By replacing zero with (ε) we will calculate all the
elements of the Routh array. After calculating all the elements, we will
apply the limit at each element containing (ε). On solving the limit at
every element if we will get positive limiting value then we will say the
given system is stable otherwise in all the other condition we will say the
given system is not stable.
(2) Case second : When all the elements of any row of the Routh array are
zero. In this case we can say the system has the symptoms of marginal
stability. Let us first understand the physical meaning of having all the
elements zero of any row. The physical meaning is that there are
symmetrically located roots of the characteristic equation in the s plane.
Now in order to find out the stability in this case we will first find out
auxiliary equation. Auxiliary equation can be formed by using the
elements of the row just above the row of zeros in the Routh array. After
finding the auxiliary equation we will differentiate the auxiliary equation to
obtain elements of the zero row. If there is no sign change in the new
routh array formed by using auxiliary equation, then in this we say the
given system is limited stable. While in all the other cases we will say the
given system is unstable.
Question bank on Routh Stability Criterion

Theory Questions
1. Explain Routh-Hurwitz criterion for determining the stability of a
system and mention is limitations.
2. What are the difficulties encountered while assessing the R- H
criterion and how do you eliminate these difficulties? Explain with
examples.
3. State and prove the theorem on Bounded Input and Bounded
Output stability.
4.