Feedback Control Systems: Dr.-Ing. Erwin Sitompul President University

Feedback Control Systems
Lecture 3
Dr.-Ing. Erwin Sitompul

President University
http://zitompul.wordpress.com
2 0 1 6
President University Erwin Sitompul FCS 3/1
Chapter 3 Dynamic Response
Effects of Zeros and Additional Poles

◼ Given a transfer function with 2 complex poles and 1 zero:
( s n ) + 1
H ( s) =
( s n ) 2 + 2 ( s n ) + 1 ζ = 0.5
◼ The zero is located at:

s = −n = −
◼ The poles are located at:
s = −  jd
◼ If  is large, the zero will be far

from the poles and the zero will
have little effect on the
response.
◼ If  ≈ 1, the value of the zero will be close to that of the real
part of the poles and can be expected to have a substantial
influence on the response.

◼ Step response of a second-order system with zero in the LHP
Overall response Original response
(having no zero)
Response introduced
by the zero
 The major effect of the zero

in the LHP is to increase the
overshoot.
 It has very little influence
on the settling time.


◼ Step response of a second-order system with zero in the RHP
Original response Overall response
(having no zero)
Response introduced
by the zero
 The major effect of the zero

in the RHP is to depress the
overshoot.
 It may cause an initial
undershoot (the step
response starts out in the
wrong direction).


◼ Given a transfer function with two complex poles and one
real pole:
1
H ( s) =
( s n + 1) ( s n ) 2 + 2 ( s n ) + 1
◼ The real pole is located at :

s = −n = −


◼ Step response of several third-order systems with ζ =0.5
ζ = 0.5
 The major effect of
an extra pole is to
increase the rise
time.
 If α is large (hence
far from the other
poles), the extra
pole have little effect
on the response.

Summary of Effects of Pole and Zero

◼ The pole and zero locations determine the character of the
transient response.
◼ A zero in the LHP will increase the overshoot if the zero is
within a factor of 4 of the real part of the complex poles ().
◼ A zero in the RHP will depress the overshoot (and may cause
the step response to start out in the wrong direction).
◼ An additional pole in the LHP will increase the rise time
significantly (slow down the response) if the extra pole is
within a factor of 4 of the real part of the complex poles.
◼ If the extra pole is more than a factor of 6 of the real part of
the complex poles, the effect is negligible.
• An additional pole in the RHP will …?

Stability
◼ Consider the linear time-invariant system (LTI system).
For those systems, the following condition for stability
applies:
“ A linear time-invariant system is said to be stable

if all the roots of the transfer function denominator
polynomial have negative real parts (i.e., they are
all in the left half of s-plane) and is unstable
otherwise.
”
<0 =0 >0
 A system is stable if its impulse

responses decay to zero and is
unstable if they diverge.
Stability of Linear Time-Invariant Systems

◼ Consider the linear time-invariant system whose transfer
function denominator polynomial leads to the characteristic
equation
s n + a1s n −1 + a2 s n − 2 + + an −1s + an = 0
◼ Assume that the roots {pi} of the characteristic equation are

real or complex, but are distinct; so that the transfer function
can be given as:
Y ( s) b0 s m + b1s m−1 + b2 s m−2 + + bm−1s + bm
T ( s) = = n
R( s ) s + a1s n−1 + a2 s n −2 + + an −1s + an
K  i =1 ( s − zi )
m
= , mn

n
i =1
( s − pi )


◼ The solution of the system response, found using partial
fraction expansion, may be written as:
n
y (t ) =  Ki e pit
i =1
◼ The system is stable if and only if (necessary and sufficient

condition) every term in the equation above goes to zero as
t → .
e pit → 0 for all pi
◼ This situation will happen if all the poles of the system are
strictly in the LHP.
Re  pi   0


◼ If any LHP poles are repeated, the response will change
because a polynomial in t must be included in place of Ki.
However, the conclusion is the same: as t → , y(t) → 0.
1
lim t n −1e− at  0 for any n ≥ 1
t → ( n − 1)!
◼ Thus, the stability of a system can be determined by

computing the location of the roots of the characteristic
equation and determining whether they are all in the LHP.
This is called internal stability.
◼ If a system has any poles in the RHP, it is unstable.
◼ If a system has non-repeated jω-axis poles, then it is said
to be neutrally stable.
◼ If the system has repeated jω-axis poles, then it is
unstable, as it results in tn–1e±jωt in the first equation of
the last slide. (Increasing multiplying factor t but no
exponential decay)
Exponential Series, Power Series

◼ Using Taylor series approximation, we can see that
exponential series eλt increases faster than power series of tk.
(  t ) 2
(  t ) 3
(  t ) 4
et = 1 + t + + + + ...
2! 3! 4!
et  t k for any value of k
◼ Therefore, e–at decreases faster than the increase of tk.

1
lim t n −1e− at  0 for any n ≥ 1
t → ( n − 1)!
◼ Also, since e  jt = cos t  j sin t then lim t n −1e  jt  0

t →
◼ Repeated jω-axis poles will make the system unstable.

Routh’s Stability Criterion

◼ The roots of the characteristic equation determine whether
the system is stable or unstable.
◼ Consider the characteristic equation

a ( s ) = s n + a1s n −1 + a2 s n − 2 + + an −1s + an
◼ Routh’s stability criterion allows us to make certain

statements about the stability of the system without actually
solving for the roots of the polynomial.
◼ Routh’s stability criterion is also useful for determining the

ranges of coefficients of polynomials for stability, especially
when the coefficients are in symbolic (non-numerical) form.


◼ A necessary condition for stability of the system is that all of
the roots of its characteristic equation have negative real
parts, which in turn requires that all the coefficients {ai} be
positive.
“ A necessary (but not sufficient) condition for

stability is that all the coefficients of the
characteristic polynomial be positive.
≡ ”
“
(identical to)
If a system is stable, then all the coefficients of
”
the characteristic polynomial are positive.
Implication p → q:
◼ If p, then q.
◼ p implies q.
◼ p only if q.
◼ p is the sufficient condition for q.
◼ q is the necessary condition for p.

◼ Once the elementary necessary conditions have been
satisfied, a more powerful test is needed.
◼ Routh in 1874 proposed a test that requires the computation
of a triangular array that is a function of the coefficients of
the characteristic equation.
“ A system is stable if and only if all the elements in

the first column of the Routh array are positive.
≡ ”
“
(identical to)
If a system is stable then all the elements in the

first column of the Routh array are positive, and
vice versa.
Bi-Implication p  q:
◼ p if and only if q.
”
◼ p is the sufficient and necessary conditions for q.
◼ if p then q, and if q then p.

◼ Consider the characteristic equation
a ( s ) = s n + a1s n −1 + a2 s n − 2 + + an −1s + an
◼ First, arrange the coefficients of the characteristic polynomial

in two rows, beginning with the first and second coefficients
and followed by the even-numbered and odd-numbered
coefficients:
sn : 1 a2 a4
s n −1 : a1 a3 a5


◼ Then add subsequent rows to complete the Routh array:
a1a2 − a3 b1a3 − b2 a1
Row n
n
s : 1 a2 a4 b1 = , c1 = ,
a1 b1
n −1
Row n − 1 s : a1 a3 a5 a1a4 − a5 b1a5 − b3 a1
n−2
b = , c = ,
Row n − 2 s
2 2
: b1 b2 b3 a1 b1
n −3 a1a6 − a7 b1a7 − b4 a1
Row n − 3 s : c1 c2 c3 b3 = , c3 = ,
a1 b1
Row 2 s2 : * *
Row 1 s1 : *
 If the elements of the first
Row 0 s : 0
* column are all positive, then
all the roots are in the LHP.
 If the elements are not all
positive, then the number of
roots in the RHP equals the
First column of number of sign changes in
Routh’s array the column.
Example 1: Routh’s Test

All the coefficients of the characteristic equation
a( s ) = s 6 + 4 s 5 + 3s 4 + 2 s 3 + 1s 2 + 4 s + 4
are positive → the system maybe stable or maybe not.
Now, we determine whether all of the roots are in the LHP
a( s ) = s 6 + 4 s 5 + 3s 4 + 2 s 3 + 1s 2 + 4 s + 4
s6 : 1 3 1 4
s5 : 4 2 4 0
s4 : ? ? ?
s3 : ? ?
s2 : ? ?
s1 : ?
s0 : ?


s6 : 1 3 1 4 43 − 2
b1 = = 2.5
s5 : 4 2 4 0 4
4 1 − 4
s 4 : 2.5
? ?0 ?4 b2 = =0
4
s3 : ? ?
44 − 0
s2 : ? ? b3 = =4
4
s1 : ?
s0 : ?


s6 : 1 3 1 4 2.5  2 − 0  4
c1 = =2
s5 : 4 2 4 0 2.5
s 4 : 2.5 0 4 2.5  4 − 4  4
c2 = = −2.4
s3 : ?2 –2.4
? 2.5
s2 : ?3 4? 2  0 − ( −2.4 ) 2.5
d1 = =3
s1 : ? 2
2  4 − 0  2.5
s0 : ? d2 = =4
2


s6 : 1 3 1 4 3  ( −2.4 ) − 4.2
e1 = = −5.067
s5 : 4 2 4 0 3
s4 : 2.5 0 4 −5.067  4 − 0.3
f1 = =4
s3 : 2 −2.4 −5.067
s2 : 3 4
s1 : −5.067
s0 : 4
• The elements of the first column are not all positive

➔ The characteristic equation has at least one
RHP root
➔ The system is unstable
• There are two sign changes (+ to – and – to +)
➔ There are two poles in the RHP


◼ Roots of polynomials can also be found by using MATLAB:
• Roots in the RHP, i.e.,

roots with positive real parts
• There are two roots of
characteristic equation in the
RHP
• There are two unstable poles


Given the characteristic equation:
a( s ) = s 6 + 4s 5 − 4s 4 + 6s 3 + 8s 2 + 1s + 1
Is the system described by this characteristic equation stable?
“ If a system is stable, then all the coefficients of the
”
characteristic polynomial are positive.
≡
“
(identical to)
If not all the coefficients of the characteristic

polynomial are positive, then
a system is not stable.
 p → q  ~q → ~p ”

a( s ) = s 6 + 4s 5 − 4s 4 + 6s 3 + 8s 2 + 1s + 1
• There is a negative coefficient
➔ The system is not stable
• Necessary condition for
stability is not even fulfilled
➔ No need to continue to
Routh’s Test
• Roots in the RHP, i.e.,

roots with positive real parts

Stability Versus Parameter Range

Consider the system shown below. The stability properties of
the system are a function of the proportional feedback gain K.
Determine the range of K over which the system is stable.
( s + 1)
K
Y ( s) s ( s − 1)( s + 6) K ( s + 1) b( s )
= = =
R( s) 1 + K ( s + 1) s( s − 1)( s + 6) + K ( s + 1) a( s )
s ( s − 1)( s + 6)
• The characteristic equation
• Which is the denominator of the
transfer function


a ( s ) = s 3 + 5s 2 + ( K − 6) s + K 5  ( K − 6) − K 4 K − 30
b1 = =
s3 : 1 ( K − 6) 5 5
b2 = 0
s2 : 5 K
b1  K − b2  5
s1 : b1 b2 c1 = =K
b1
s 0 : c1 c2 c2 = 0
The system is stable if and only if b1 and c1 are positive.

4 K − 30  0  K  7.5 
  K  7.5
K 0 


Generating the step responses of the transfer function in
MATLAB, for 3 different values of K:
Y (s) Ks + K
= 3
R( s) s + 5s 2 + ( K − 6)s + K
10
5
Amplitude
-5
-10 K=5
K=7.5
K=10
-15
0 5 Time (sec) 10

Stability Versus Two Parameter Ranges

Find the range of the controller gains (K, KI) so that the PI
(proportional-integral) feedback system in the figure below is
stable.
 KI  1
 K+ 
Y ( s)  s  ( s + 1)( s + 2) Ks + K I
= = 3
R( s)  KI  1 s + 3s 2 + (2 + K ) s + K I
1+  K + 
 s  ( s + 1)( s + 2) The characteristic equation

Stability Versus Two Parameter Ranges

a( s ) = s 3 + 3s 2 + (2 + K ) s + K I
s3 : 1 2+ K K
s2 : 3 KI
1 3  (2 + K ) − K I allowable
s : 0 region
3
s0 : KI 0
K
For the stability, I
6 + 3K − K I  0 
 K I  0
KI  0 
KI − 6
K
3
Special Cases for Routh’s Stability Criterion

◼ If a first-column term in any row is zero, but the remaining
terms are not zero or there is no remaining term, then the
zero term is replaced by a very small positive number ε and
the rest of the array is evaluated as before.
Example:
a( s) = s3 + 2s 2 + s + 2
s3 : 1 1
s2 : 2 2
s1 : 0   0
s0 : 2 0


◼ If the sign of the coefficient above the zero (ε) is the same
as that below it, it indicates that there are a pair of
imaginary roots.
Example:
a( s) = s3 + 2s 2 + s + 2
s3 : + 1 1
s2 : + 2 2
s1 : + 0   0
s1 = −2
s :+ 2
0
0 s2,3 =  j
The same signs above and below zero (ε)
➔ A pair of imaginary roots


◼ If the sign of the coefficient above the zero (ε) is opposite
that below it, it indicates that there is one sign change.
Example:
a ( s ) = s 5 + 3s 4 + 2s 3 + 6s 2 + 6s + 9
s5 : + 1 1st 2 6
s4 : + 3 6 9 The opposite signs above and below ε
s3 : + 0 3 ➔ There is one sign change
➔ The 1st root in the RHP
s2 : − 6 −9 9
 Another sign change between s2 and s1
2nd ➔ The 2nd root in the RHP
s1 : +
3  (
6 −9 − 9
 )
6 −9 s1 = −2.9043

s2,3 = 0.6567  j1.2881
s0 : + 9
s4,5 = −0.7046  j 0.9929

◼ If all the coefficients in any derived row are zero, it indicates
that there are roots of equal magnitude lying radially
opposite in the s-plane, that is, two real roots with equal
magnitudes and opposite signs and/or two conjugate
imaginary roots.
◼ In such a case, the evaluation of the rest of the array can be

continued by forming an auxiliary polynomial from the last
nonzero row, and then using the coefficients of the derivative
of this auxiliary polynomial to replace the zero row.


Example:
a( s ) = s 5 + 5s 4 + 11s 3 + 23s 2 + 28s + 12
s5 : 1 11 28
s4 : 5 23 12
s 3 : 6.4 25.6 0 Auxiliary polynomial from
the last nonzero row
s2 : 3 12  a1 ( s ) = 3s 2 + 12
s1 : 0 0 Zero row
da1 ( s ) Derivative of the auxiliary
 = 6s
1
New s : 6 0 polynomial
ds
s 0 : 12 Zero row is s1 = −3
replaced
s2,3 =  j 2
• One zero row
➔ Radially opposite roots s4 = −1
• No sign change
➔ No root in the RHP s5 = −1

Example: Unstable
a( s ) = s 5 + 2 s 4 + 24s 3 + 48s 2 − 25s − 50

s5 : 1 24 −25 Auxiliary polynomial from
the last nonzero row
s4 : 2 48 −50  a1 ( s ) = 2s 4 + 48s 2 − 50
Zero row
s3 : 0 0 0
da1 ( s )
New
3
s : 8 96 0  = 8s 3 + 96s
ds Derivative of the
s2 : 24 −50 auxiliary polynomial
s1 : 112.7 0 Zero row is

s0 : −50
replaced s1,2 =  j5
s3 =1
• One zero row s4 = −1
➔ Radially opposite roots
• One sign change s5 = −2
➔ One root in the RHP

Homework 3
◼ No.1
A system is given with a transfer function:
Y (s) 24( s + 1)
= 4
R( s ) s + 6s 3 + 2s 2 + s + 3
Determine whether the system is stable or not. Use Routh’s
stability criterion and verify with MATLAB.
◼ No.2
A system has a characteristic equation:
s 3 + 2 s 2 + ( K + 1) s + 6 = 0
Find the range of K for a stable system.
◼Deadline: Wednesday, 28 September 2016.

◼Quiz 1: Thursday, 29 September 2016
from Lecture 1-3.

Feedback Control Systems: Dr.-Ing. Erwin Sitompul President University

Uploaded by

Copyright:

Available Formats

You might also like

Feedback Control Systems: Dr.-Ing. Erwin Sitompul President University

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Feedback Control Systems: Dr.-Ing. Erwin Sitompul President University

Uploaded by

Copyright:

Available Formats

Feedback Control Systems

Dr.-Ing. Erwin Sitompul

Effects of Zeros and Additional Poles

◼ The zero is located at:

◼ If  is large, the zero will be far

Effects of Zeros and Additional Poles

 The major effect of the zero

President University Erwin Sitompul FCS 3/3

Effects of Zeros and Additional Poles

 The major effect of the zero

President University Erwin Sitompul FCS 3/4

Effects of Zeros and Additional Poles

◼ The real pole is located at :

President University Erwin Sitompul FCS 3/5

Effects of Zeros and Additional Poles

President University Erwin Sitompul FCS 3/6

Summary of Effects of Pole and Zero

• An additional pole in the RHP will …?

President University Erwin Sitompul FCS 3/7

“ A linear time-invariant system is said to be stable

 A system is stable if its impulse

Stability of Linear Time-Invariant Systems

◼ Assume that the roots {pi} of the characteristic equation are

President University Erwin Sitompul FCS 3/9

Stability of Linear Time-Invariant Systems

◼ The system is stable if and only if (necessary and sufficient

e pit → 0 for all pi

President University Erwin Sitompul FCS 3/10

Stability of Linear Time-Invariant Systems

◼ Thus, the stability of a system can be determined by

Exponential Series, Power Series

◼ Therefore, e–at decreases faster than the increase of tk.

◼ Also, since e  jt = cos t  j sin t then lim t n −1e  jt  0

◼ Repeated jω-axis poles will make the system unstable.

President University Erwin Sitompul FCS 3/12

Routh’s Stability Criterion

◼ Consider the characteristic equation

◼ Routh’s stability criterion allows us to make certain

◼ Routh’s stability criterion is also useful for determining the

President University Erwin Sitompul FCS 3/13

Routh’s Stability Criterion

“ A necessary (but not sufficient) condition for

If a system is stable, then all the coefficients of

Routh’s Stability Criterion

“ A system is stable if and only if all the elements in

If a system is stable then all the elements in the

Routh’s Stability Criterion

◼ First, arrange the coefficients of the characteristic polynomial

President University Erwin Sitompul FCS 3/16

Routh’s Stability Criterion

Example 1: Routh’s Test

President University Erwin Sitompul FCS 3/18

Example 1: Routh’s Test

President University Erwin Sitompul FCS 3/19

Example 1: Routh’s Test

President University Erwin Sitompul FCS 3/20

Example 1: Routh’s Test