Вы находитесь на странице: 1из 48

Chapter 5

THE ROOT-LOCUS DESIGN METHOD

Feedback Control System


Dr.-Ing. Erwin Sitompul
http://zitompul.wordpress.com
Root Locus: Illustrative Example
Examine the following closed-loop system, with unity
negative feedback.

The closed-loop transfer The roots of the characteristic


function is given as: equation are:
Y ( s)
2
K 1 1 4 K
s1,2
R( s ) s s K 2

The characteristic equation


The denominator of the
closed-loop transfer function

Erwin Sitompul Feedback Control System 6/2


Root Locus: Illustrative Example
1 1 4K 1
, 0 K

s1,2 2 2 4
1 j 4 K 1 , K 1
2 2 4

2
K=
: the poles of open-loop
K=1/4 transfer function
1
Imaginary Axis

0 K=0 K=0

-1

K=
-2
-2 -1 0 1
Real Axis
Erwin Sitompul Feedback Control System 6/3
Root Locus: Illustrative Example
Where are the location of the closed-loop roots
when K=1? n2 1 n 1
Y ( s) K 1 2n 1 0.5

R( s ) s2 s K s2 s 1
n 0.5
1.5
d n 1 2 0.866
1 K=1
0.866
Imaginary Axis

0.5
( jd ) (0.5 j 0.866)
0
There is a relation
-0.5 between gain K and the
0.866 position of closed-loop
-1 K=1 poles, which also affects
the dynamic properties of
1 the system ( and d)
-1.5
-2 -1 0.5 0
Real Axis
Erwin Sitompul Feedback Control System 6/4
Root Locus of a Basic Feedback System

The closed-loop transfer function of the basic feedback


system above is:
Y (s) D( s)G ( s )
T ( s)
R( s ) 1 D( s)G ( s) H ( s )

The characteristic equation, whose roots are the poles of this


transfer function, is:
1 D( s)G ( s) H ( s) 0
Erwin Sitompul Feedback Control System 6/5
Root Locus of a Basic Feedback System
To put the characteristic equation in a form suitable for study
of the roots as a parameter changes, it is rewritten as:
1 KL( s ) 0
where
KL( s) D( s)G ( s) H ( s)
b( s )
L( s )
a( s)
K is the gain of controller-plant-sensor combination.
K is selected as the parameter of interest.
W. R. Evans (in 1948, at the age of 28) suggested to plot
the locus (location) of all possible roots of the characteristic
equation as K varies from zero to infinity root locus plot.
The resulting plot is to be used as an aid in selecting the
best value of K.
Erwin Sitompul Feedback Control System 6/6
Root Locus of a Basic Feedback System
The root locus problem shall now be expressed in several
equivalent but useful ways.
1 KL( s ) 0
b( s )
1 K 0
a( s)
a( s ) Kb( s ) 0
1
L( s )
K
The equations above are sometimes referred to as the
root locus form of a characteristic equation.
The root locus is the set of values of s for which the above
equations hold for some positive real value of K.

Erwin Sitompul Feedback Control System 6/7


Root Locus of a Basic Feedback System
Explicit solutions are difficult to obtain for higher-order
system General rules for the construction of a root locus
were developed by Evans.
With the availability of MATLAB, plotting a root locus
becomes very easy, using the command rlocus(num,den).
However, in control design we are also interested in how to
modify the dynamic response so that a system can meet
the specifications for good control performance.
For this purpose, it is very useful to be able to roughly
sketch a root locus which will be used to examine a system
and to evaluate the consequences of possible
compensation alternatives.
Also, it is important to be able to quickly evaluate the
correctness of a MATLAB-generated locus to verify that
what is plotted is in fact what was meant to be plotted.

Erwin Sitompul Feedback Control System 6/8


Guideines for Sketching a Root Locus
Deriving using the root locus form of characteristic equation,
1 KL( s ) 0
1
L( s )
K
b( s ) 1

a( s) K
Taking the polynomial a(s) and b(s) to be monic, i.e., the
coefficient of the highest power of s equals1, they can be
factorized as:
( s z1 )( s z2 ) ( s zm ) 1

( s p1 )( s p2 ) ( s pn ) K
If any s = s0 fulfills the equation above, then s0 is said to be
on the root locus.

Erwin Sitompul Feedback Control System 6/9


Guideines for Sketching a Root Locus
The magnitude condition implies:
( s z1 ) ( s z2 ) ( s zm ) 1 1
Magnitude
( s p1 ) ( s p2 ) ( s pn ) s s0
K K Condition

The phase condition implies:


( s z1 )( s z2 ) ( s zm ) 1
180 Phase
( s p1 )(s p2 ) (s pn ) s s K Condition
0

Defining (s zi ) i and ( s pi ) i , the phase


condition can be rewritten as:
m n


i 1
i
i 1
i 180 360(l 1), l 1, 2,3,

Erwin Sitompul Feedback Control System 6/10


Guideines for Sketching a Root Locus

The root locus is the set of values of s for which


1 + KL(s) = 0 is satisfied as the real parameter K varies
from 0 to . Typically, 1 + KL(s) = 0 is the characteristic
equation of the system, and in this case the roots on the


locus are the closed-loop poles of that system.

The root locus of L(s) is the set of points in the s-plane


where the phase of L(s) is 180. If the angle to a test
point from a zero is defined as i and the angle to a
test point from a pole as i, then the root locus of L(s)
is expressed as those points in the s-plane where, for


integer l, i i = 180 + 360(l1).

Erwin Sitompul Feedback Control System 6/11


Guideines for Sketching a Root Locus
Consider the following example.
s 1
L( s )
s(s 5) ( s 2)2 4
: the poles of L(s)
: the zero of L(s)

s0 1 j 2 Test
point
1 90
1 tan 1 (2 1) 116.6
i i 1 1 2 3 4
2 0
90 116.6 0 76.0 26.6
3 tan 1 (4 1) 76.0
129.2 180 360(l 1)
4 tan 1 (2 4) 26.6 s0 is not on the root locus
Erwin Sitompul Feedback Control System 6/12
Rules for Plotting a Root Locus
RULE 1:
The n branches of the locus start at the poles of L(s) and m of
these branches end on the zeros of L(s), while nm branches
terminate at infinity along asymptotes.

Recollecting
b( s ) 1 b( s ) 1 ( s z1 ) ( s z2 ) ( s zm ) 1

a( s) K a( s) K ( s p1 ) ( s p2 ) ( s pn ) K
1 ( s z1 ) ( s z2 ) ( s zm )
lim lim
K 0 K s pi ( s p ) ( s p ) ( s pn )
1 2

1 ( s z1 ) ( s z2 ) ( s zm )
lim 0 lim
K K s zi ( s p ) ( s p ) ( s pn )
1 2

The root locus starts at K = 0 at the poles of L(s)


and ends at K = on the zeros of L(s)
Erwin Sitompul Feedback Control System 6/13
Rules for Plotting a Root Locus
RULE 2:
On the real axis, the loci (plural of locus) are to the left of an
odd number of poles and zeros.
5 4 3 2 1
: The root locus

4 3 2 1

Angles from real poles or zeros are 0 if the test point is


to the right and 180 if the test point is to the left of a
given pole or zero.
Erwin Sitompul Feedback Control System 6/14
Rules for Plotting a Root Locus
The rule is now applied to obtain the root locus of:
1
L( s ) p1 0
s (s 4)2 16 p2,3 4 j 4

Erwin Sitompul Feedback Control System 6/15


Rules for Plotting a Root Locus
For any test point s0 on the real 1 tan 1 (2 4) 26.6
axis, the angles 1 and 2 of two
complex conjugate poles cancel 2 tan 1 (2 4) 26.6
each other, as would the angles of 1 2 0
two complex conjugate zeros
(see figure below). 3 0 The pair does not
give contribution to
the phase condition

1 2 3 0
s1
180 360(l 1)
s0 is not on the
root locus

Now, check the phase


condition of s1!

Erwin Sitompul Feedback Control System 6/16


Rules for Plotting a Root Locus
RULE 3:
For large K and s, nm of the loci are asymptotic to lines at
angles l radiating out from the point s = on the real axis,
where:
180 360(l 1)
l , l 1, 2, ,3 Angles of Asymptotes
nm

pi zi
Center of Asymptotes
nm

Erwin Sitompul Feedback Control System 6/17


Rules for Plotting a Root Locus
1
For L(s) , we obtain n 3, m 0
s (s 4) 16
2
p1 0, p2,3 4 j 4
180 360(l 1) 60
l
nm
180 360(l 1)

30
60 120(l 1) 180
60,180,300 2.67

p z
i i

nm
(4 j 4) (4 j 4) 0
300
30
8
2.67
3
Erwin Sitompul Feedback Control System 6/18
Rules for Plotting a Root Locus
RULE 4:
The angle of departure of a branch of a locus from a pole is
given by:
l ,dep i i 180 360(l 1)
i l

and the angle of arrival of a branch of a locus to a zero is


given by:
l ,arr i i 180 360(l 1)
i l

Erwin Sitompul Feedback Control System 6/19


Rules for Plotting a Root Locus
For the example, the root loci must depart with certain angles
from the complex conjugate poles at 4 j4, and go to the
zero at with the angles of asymptotes 60 and 300.

Erwin Sitompul Feedback Control System 6/20


Rules for Plotting a Root Locus
From the figure,
2 i i 180 360(l 1)
i2
2 1 3 180 360(l 1)
But
1 90
3 tan 1 (4 4) 135
Thus
2 90 135 180
2 405 45

By the complex conjugate symmetry


of the roots, the angle of departure
of the locus from 4 j4 will be +45.

Erwin Sitompul Feedback Control System 6/21


Rules for Plotting a Root Locus
So, the root loci will start their journey from 4 j4 towards

with the direction of 45.

Erwin Sitompul Feedback Control System 6/22


Rules for Plotting a Root Locus
RULE 5:
The locus crosses the j axis (imaginary axis) at points where:
The Routh criterion shows a transition from roots in the left
half-plane to roots in the right half-plane.
This transition means that the closed-loop system is
becoming unstable.
This fact can be tested by Rouths stability criterion, with
K as the parameter, where an incremental change of K
will cause the sign change of an element in the first
column of Rouths array.
The values of s = j0 are the solution of the characteristic
equation in root locus form, 1 + KL(s) = 0.
The points j0 are the points of cross-over on the
imaginary axis.

Erwin Sitompul Feedback Control System 6/23


Rules for Plotting a Root Locus
For the example, the characteristic equation can be written as:
1 KL( s ) 0
1
1 K 0
s ( s 4) 16
2

s 3 8s 2 32s K 0

s3 : 1 32
s2 : 8 K The closed-loop system is stable for K > 0
and K < 256 for 0 < K < 256.
1 8 32 K For K > 256 there are 2 roots in the RHP
s : (two sign changes in the first column).
8 For K = 256 the roots must be on the
imaginary axis.
s0 : K

Erwin Sitompul Feedback Control System 6/24


Rules for Plotting a Root Locus
The characteristic equation is now solved using K = 256.
s 3 8s 2 32s 256 0
s1 8
s2,3 j 5.66 j0 Points of Cross-over

Another way to solve for 0 is by simply replacing any s with j0


without finding the value of K first.
( j0 )3 8( j0 ) 2 32( j0 ) K 0
j03 80 2 j 320 K 0
K 80 2 j (320 03 ) 0 320 03 K 80 2
0 2 32 K 8 32
0 0 0 5.66 K 256
Same results for K and 0
Erwin Sitompul Feedback Control System 6/25
Rules for Plotting a Root Locus
The points of cross-over are now inserted to the plot.
5.66

5.66

Erwin Sitompul Feedback Control System 6/26


Rules for Plotting a Root Locus
The complete root locus plot can be shown as:

Erwin Sitompul Feedback Control System 6/27


Rules for Plotting a Root Locus
RULE 6:
The locus will have multiple roots at points on the locus where:
da( s ) db( s )
b( s ) a( s) 0
ds ds
The branches will approach and depart a point of q roots at
angles separated by:
180 360(l 1)
q

Erwin Sitompul Feedback Control System 6/28


Rules for Plotting a Root Locus
A special case of point of multiple roots is the intersection
point of 2 roots that lies on the real axis.
If the branches is leaving the real axis and entering the
complex plane, the point is called the break-away point.
If the branches is leaving the complex plane and entering
the real axis, the point is called the break-in point.
Imag Imag
axis axis

90 90
Real Real
axis axis
Break-away Break-in
point point

Erwin Sitompul Feedback Control System 6/29


Example 1: Plotting a Root Locus
Draw the root locus plot of the system shown below.

K
Y ( s) s ( s 1)( s 2) 1
L( s )
R( s) 1 K s( s 1)( s 2) RULE 1
s ( s 1)( s 2) n 3, m 0 3 zeros at infinity
p1 0, p2 1, p3 2
1 KL( s )

Erwin Sitompul Feedback Control System 6/30


Example 1: Plotting a Root Locus
Imag
axis
p1 0, p2 1, p3 2 3

RULE 2 1

Real
4 3 2 1 0 1 2 axis

Erwin Sitompul Feedback Control System 6/31


Example 1: Plotting a Root Locus
180 360(l 1)
l
nm
180 360(l 1)

30
60 120(l 1)
60,180,300 Angles of Asymptotes

p z i i
p1 0, p2 1, p3 2
nm
0 1 2

30
1 Center of Asymptotes

Erwin Sitompul Feedback Control System 6/32


Example 1: Plotting a Root Locus
Imag
axis
l 60,180,300 3
60
1 2

180 Real
4 3 2 1 0 1 2 axis

1 RULE 3
RULE 4
2
Not applicable. The angles 300
of departure or the angles of
arrival must be calculated 3
only if there are any complex
poles or zeros.
Erwin Sitompul Feedback Control System 6/33
Example 1: Plotting a Root Locus
1 KL( s ) 0
1
1 K 0
s( s 1)( s 2)
s 3 3s 2 2s K 0

Replacing s with j0,


( j0 )3 3( j0 ) 2 2( j0 ) K 0
j03 30 2 j 20 K 0
K 30 2 j (20 03 ) 0 20 03 K 30 2
0 2 2 K 3 2
0 0 0 1.414 K 6
Points of Cross-over

Erwin Sitompul Feedback Control System 6/34


Example 1: Plotting a Root Locus
Imag
3 axis
0 1.414
2
1.414 RULE 5
1

Real
4 3 2 1 0 1 2 axis

1
1.414
2

Erwin Sitompul Feedback Control System 6/35


Example 1: Plotting a Root Locus
1 KL( s ) 0
b( s ) 1 1
L( s ) 3
a( s) s( s 1)( s 2) s 3s 2 2s
The root locus must have a break-away point, which
can be found by solving:
da( s ) db( s )
b( s ) a( s) 0
ds ds
1 (3s 2 6s 2) (s3 3s 2 2s) 0 0
3s 2 6s 2 0
s1 1.577, s2 0.423

Erwin Sitompul Feedback Control System 6/36


Example 1: Plotting a Root Locus
Imag
s1 1.577, s2 0.423 axis
3

Not on the On the root locus


root locus The break-away 2
point 1.414
0.423 RULE 6
1

Real
4 3 2 1 0 1 2 axis

1
1.414
2

Erwin Sitompul Feedback Control System 6/37


Example 1: Plotting a Root Locus
After examining RULE 1 up to Imag
RULE 6, now there is enough 3 axis
information to draw the root
locus plot.
2
1.414
0.423
1
90
Real
4 3 2 1 0 1 2 axis

1
1.414
2

Erwin Sitompul Feedback Control System 6/38


Example 1: Plotting a Root Locus
The final sketch, with direction of Imag
root movements as K increases 3 axis
from 0 to can be shown as:
2 Final Result
1.414
0.423
1

Real
4 3 2 1 0 1 2 axis

1
1.414
2

3
Determine the locus of
all roots when K = 6!
Erwin Sitompul Feedback Control System 6/39
Example 2: Plotting a Root Locus

a) Draw the root locus plot of the system.


b) Define the value of K where the system is stable.
c) Find the value of K so that the system has a root at s = 2.

Erwin Sitompul Feedback Control System 6/40


Example 2: Plotting a Root Locus

1
K
Y ( s) ( s 4) s2
L( s ) RULE 1
R( s ) 1 K 1 ( s 2) s 4 There is one branch,
( s 4) starts from the pole and
n 1, m 1 approaches the zero
p1 4, z1 2
1 KL( s )

Erwin Sitompul Feedback Control System 6/41


Example 2: Plotting a Root Locus
Imag
axis
p1 4, z1 2 3

RULE 2 1

Real
RULE 3 4 3 2 1 0 1 2 axis
Not applicable, since n = m. 1

RULE 4
2
Not applicable. The angles
of departure or the angles of
arrival must be calculated 3
only if there are any complex
poles or zeros.
Erwin Sitompul Feedback Control System 6/42
Example 2: Plotting a Root Locus
1 KL( s ) 0
s2
1 K 0
s4
s 4 K ( s 2) 0

Replacing s with j0,


( j0 ) 4 K ( j0 2) 0
4 2K j0 (1 K ) 0 4 2K 0 (1 K )
K 2 0 0
0 0
Points of Cross-over

Erwin Sitompul Feedback Control System 6/43


Example 2: Plotting a Root Locus
Imag
RULE 5 axis
3
The point of cross-over, as can
readily be guessed, is at s = 0.
2

K=2
1

Real
4 3 2 1 0 1 2 axis

1
RULE 6
2
Not applicable. There is no
break-in or break-away point.
3

Erwin Sitompul Feedback Control System 6/44


Example 2: Plotting a Root Locus
a) Draw the root locus plot of the system. Imag
3 axis
The final sketch, with direction of
root movements as K increases Final Result
from 0 to can be shown as: 2

Real
4 3 2 1 0 1 2 axis

Erwin Sitompul Feedback Control System 6/45


Example 2: Plotting a Root Locus
b) Define the value of K where the Imag
system is stable. 3 axis

K=0 K=2 K=
1

Real
4 3 2 1 0 1 2 axis

1
System is stable when the root of
the characteristic equation is on 2
the LHP, that is when 0 K < 2.
3

Erwin Sitompul Feedback Control System 6/46


Example 2: Plotting a Root Locus
c) Find the value of K so that the Imag
system has a root at s = 2. 3 axis

K = 0.5 2

K=0 K=2 K=
1

Real
4 3 2 1 0 1 2 axis

Inserting the value of s = 2 in 1


the characteristic equation,
2
s 4 K ( s 2) s 2 0
2 4 K (2 2) 0 3
K (2 2) 2 4 0.5

Erwin Sitompul Feedback Control System 6/47


Homework 6
No.1, FPE (5th Ed.), 5.2.
Hint: Easier way is to assign reasonable values for the zeros and poles in
each figure. Later, use MATLAB to draw the root locus.

No.2, FPE (5th Ed.), 5.7.(b)


Hint: After completing the hand sketch, verify your result using MATLAB.
Try to play around with Data Cursor.
No.3
Sketch the root locus diagram of the following closed-
loop system as accurate as possible.

Erwin Sitompul Feedback Control System 6/48

Вам также может понравиться