You are on page 1of 41

KE36303 CONTROL SYSTEMS

Assoc. Prof Ir Dr Yang Soo Siang


BEng(Hons) MSc PhD MIEM PEng
Room 28 Level 3 Block A
Faculty of Engineering
Universiti Malaysia Sabah
KE36303 CONTROL SYSTEMS

LECTURE 5: STABILITY ANALYSIS

Assoc Prof Ir Dr Yang Soo Siang 2


OVERVIEW
General.

Routh-Hurwitz criterion.

Stability design.

Assoc Prof Ir Dr Yang Soo Siang 3


GENERAL
Transient requirements: time constant, rise
time, settling time, peak overshoot,
damping ratio etc

Steady state requirements: errors

Stability: MOST IMPORTANT SYSTEM


SPECIFICATION! WHY?
Assoc Prof Ir Dr Yang Soo Siang 4
GENERAL
If system is unstable need not consider other
specifications no basis for controller design.

Formal definitions:
An LTI system is stable if the natural response approaches zero as time
approaches infinity.

An LTI system is unstable if the natural response grows without bound


as time approaches infinity.

An LTI system is marginally stable if the natural response neither


decays nor grows but remains constant or oscillates as time approaches
infinity.
Assoc Prof Ir Dr Yang Soo Siang 5
GENERAL
OR can be described as:
A system is stable if every bounded input
yields a bounded output

A system is unstable if any bounded input


yields an unbounded output

this definition is more relevant in terms of control system


stability!
Assoc Prof Ir Dr Yang Soo Siang 6
GENERAL
Remember which part of the transfer function effects
system stability?

Hint:
if poles in left half plane (s-plane).

If poles in right half plane (s-plane).

Assoc Prof Ir Dr Yang Soo Siang 7


GENERAL

Assoc Prof Ir Dr Yang Soo Siang 8


GENERAL
Unstable systems- physically results in
damage of equipment, adjacent properties
and most importantly human lives.

Systems are designed with limit stops to


prevent total runaway.

Assoc Prof Ir Dr Yang Soo Siang 9


GENERAL
Not all mathematical models or transfer
functions are easily factorised for you to
observe their poles conveniently!

For example l
ant f
ev y o
re ilit
it ll lab n?
s ai tio
e
u av isa
s
is t or
his e fac
t th r
is fo
e
ith war
w oft
s

Assoc Prof Ir Dr Yang Soo Siang 10


ROUTH-HURWITZ CRITERION

This method- yields stability information without


the need to solve for the closed loop poles.

Results- the number of closed loop system poles


in the left half plane, in the right half plane and
on the j axis.

How many but not where! So back to the previous questions why
is this method still relevant? Hint: for controller design able to yield
a range of parameters to ensure stability of system.

Assoc Prof Ir Dr Yang Soo Siang 11


ROUTH-HURWITZ CRITERION

Requires two steps:

Generate a data table known as a Routh table

Interpret the Routh table to know how many


closed loop poles are in the left half plane
etc...

Assoc Prof Ir Dr Yang Soo Siang 12


ROUTH-HURWITZ CRITERION

Generating the Routh table:

For example:

Begin by labeling the rows with powers of s from the highest of the
denominator of the closed loop transfer function to s0

Assoc Prof Ir Dr Yang Soo Siang 13


ROUTH-HURWITZ CRITERION

Start with coeff of the highest power of


s in the denominator and list
horizontally in the first row, every other
coeff.

Assoc Prof Ir Dr Yang Soo Siang 14


ROUTH-HURWITZ CRITERION

In the second row, list horizontally


starting with the next highest power of
s, every coeff that was skipped in the
first row.

Assoc Prof Ir Dr Yang Soo Siang 15


ROUTH-HURWITZ CRITERION

Generating the Routh table: the remaining


entry
Each entry is a negative determinant of entries in the previous
two rows divided by the entry in the first columns directly above
the calculated row.

The left hand column of the determinant is always the first


column of the previous two rows, and the right hand column is
the elements of the column above and to the right.

The table is complete when all the rows are completed down to
s0.

Assoc Prof Dr Ir Yang Soo Siang 16


ROUTH-HURWITZ CRITERION

Generating the Routh table:

Assoc Prof Ir Dr Yang Soo Siang 17


ROUTH-HURWITZ CRITERION

For example:

Assoc Prof Ir Dr Yang Soo Siang 18


ROUTH-HURWITZ CRITERION

for convenience any row can be multiplied by a positive constant


without changing the value of the rows below. Not to be multiplied by
negative constants!

Assoc Prof Ir Dr Yang Soo Siang 19


ROUTH-HURWITZ CRITERION

Interpretation:
The Routh-Hurwitz criterion declares that the number
of roots of the polynomial that are in the right half
plane is equal to the number of sign changes in the
first column.

If closed loop tf has all poles in LHP then system is


stable; no sign change in the first column!

From the example shown two poles in right RHP


Why? (based on Routh-Hurwitz criterion)
Assoc Prof Ir Dr Yang Soo Siang 20
ROUTH-HURWITZ CRITERION

Positive number

Negative number

Positive number

2 sign change = 2 RHP poles exist! Hence system is unstable!


unstable

Assoc Prof Ir Dr Yang Soo Siang 21


ROUTH-HURWITZ CRITERION

Special cases:

Zero in first column of a row

Entire row consisting of zero

Assoc Prof Ir Dr Yang Soo Siang 22


ROUTH-HURWITZ CRITERION

For example (zero in first column):

T(s)= 10/s5+2s4+3s3+6s2+5s+3

Assoc Prof Ir Dr Yang Soo Siang 23


ROUTH-HURWITZ CRITERION

Assoc Prof Ir Dr Yang Soo Siang 24


ROUTH-HURWITZ CRITERION

Replace zero by a small number, .

Assume a sign, positive or negative for the


quantity .

Assoc Prof Ir Dr Yang Soo Siang 25


ROUTH-HURWITZ CRITERION

Assoc Prof Ir Dr Yang Soo Siang 26


ROUTH-HURWITZ CRITERION

Whether positive or negative, results of


interpretation will be the same.

For the example, system is unstable with 2


poles in RHP.

Assoc Prof Ir Dr Yang Soo Siang 27


ROUTH-HURWITZ CRITERION

Alternatively

Write a polynomial that has reciprocal


roots of the denominator write the
denominator in reverse order,

Assoc Prof Ir Dr Yang Soo Siang 28


ROUTH-HURWITZ CRITERION

Via the same example:

T(s)= 10/s5+2s4+3s3+6s2+5s+3

The denominator in reverse order

D(s)= 3s5+5s4+6s3+3s2+2s+1

Assoc Prof Ir Dr Yang Soo Siang 29


ROUTH-HURWITZ CRITERION

Two sign changes hence system


is unstable and has two RHP
poles!

Assoc Prof Ir Dr Yang Soo Siang 30


ROUTH-HURWITZ CRITERION

For example (row of zeros*):

T(s)= 10/s5+7s4+6s3+42s2+8s+56

* zero in magnitude NOT zero of transfer function!

Assoc Prof Ir Dr Yang Soo Siang 31


ROUTH-HURWITZ CRITERION

Assoc Prof Ir Dr Yang Soo Siang 32


ROUTH-HURWITZ CRITERION

Second row divided by 7 for convenience.

Third row all zeros hence

Observe the row immediately above the row of zeros,


use entries in that row for coeff to form polynomial to
replace all zeros in the 3rd row.

P(s)=s4+6s2+8

Differentiating, dP(s)/ds=4s3+12s+0
Assoc Prof Ir Dr Yang Soo Siang 33
ROUTH-HURWITZ CRITERION

Use the coeff from the differentiated polynomial to replace the zeros,

4s3+12s+0

For convenience, multiplied divide by 4 after replacing the zeros.

Remainder of row is formed in a straightforward manner by following


the standard form.

Obviously, there are no RHP poles!

Assoc Prof Ir Dr Yang Soo Siang 34


STABILITY DESIGN

Stability design example: find the range of gain K for


system to be stable, unstable and marginally stable.
Assume K>0.

Find the closed loop transfer function.


Form the Routh table

Assoc Prof Ir Dr Yang Soo Siang 35


STABILITY DESIGN

Stability design example:

Assoc Prof Ir Dr Yang Soo Siang 36


STABILITY DESIGN

Stability design example:

Since K is assumed positive, we see all elements in the first


column are always positive except for the s1 row.

This entry can be positive, negative or zero.

If K < 1386 all terms will be positive, hence no sign change- 3


poles on the LHP and stable.

If K > 1386 the s1 term will be negative, hence 2 sign change-


2 poles on the RHP and 1 pole in LHP, system unstable.

Assoc Prof Ir Dr Yang Soo Siang 37


STABILITY DESIGN

Stability design example:

If K = 1386 we have entire row of zero

Returning to the s2 row and replacing K with 1386, form the


polynomial,
P(s)=18s2+1386

Differentiating with respect to s,


dP(s)/ds = 36s+0

hence replace coeff in the row of zeros


Assoc Prof Ir Dr Yang Soo Siang 38
STABILITY DESIGN

Stability design example:

Assoc Prof Ir Dr Yang Soo Siang 39


STABILITY DESIGN

Stability design example:

No change of sign hence the even polynomial down


to bottom of table.

Even polynomial has two roots on the jw axis.

No sign change above even polynomial, hence


remaining root is in LHP-  system is marginally
stable.

Assoc Prof Ir Dr Yang Soo Siang 40


NEXT LECTURE
Root Locus:
Basic stuff- significance etc.
Plotting and sketching

What you need to do!


Review this lecture and try out examples in Chp 5, Nise till pg
305. They are all relevant for your understanding and for you to
be familiar with forming the Routh table and stability design via
Routh Hurwitz.

In addition, read about root locus of course!

Assoc Prof Ir Dr Yang Soo Siang 41