Solution 5.8.3.3
For the system of Figure 1 let
G
H
+R
C
Figure 1: Standard Closed Loop Conguration
GH =
K(s +2)
(s +1;j3)(s+1+j3)(s+15)(s+20)
=
The rst step in drawing the root locus is to plot the poles and zeros of
GH.Thepoles of GH are not the closed loop pole locations, but they can
be used to nd the closed loop poles. The closed loop zeros can be found
immediately: they are the zeros of G and the poles of H.Thepoles and
zeros of GH serves as landmarks that help in nding the poles of the closed
loop system.
The portion of the root locus on the real axis is the shaded regions shown
in Figure 2. These regions are determined byinvoking the rule that states
that root locus on the real axis is found to the left of an odd countofpoles
and zeros of GH.
The root locus has four poles and one nite zero. One of the limbs of
the root locus will end at the nite zero for K = 1.Theother three limbs
Im(s)
Re(s)
X
X
XX
-3-15-20
Figure 2: Root Locus on Real Axis
1
will end at so-called `zeros at innity.' These zeros are at the `end' of each
of the three asymptotes. The asymptotes are at =60
;;180
;; and 300
.
The asymptotes intersect at
i
=
[0 + (;1) + (;1)+(;15) + (;20)];[;3]
3
= ;12:3:
The twopotential root loci are shown in Figure 3. If a break-in and
break-out pointexistbetween s = ;3ands = ;15 then the root locus will
look likepart (a) of the gure. If there are no break-in and break-out points
the root locus will look likepart(b) of the gure.
Tocheckforthe possibilityofbreak-in and break-out points the gain can
be calculated at some representativevaules along the real axis. The plot will
look likepart (a) of Figure 4 if there are break-in and break-out points, and
like part(b) if there are no break-in or break-out points. Figure 4.
The gain can be plotted using Figure 5. A representativepointonthe
line segmentisshown in the gure. For this choice of s we know that
jGH(s)j=
Kjs +3j
js +1;j3jjs+1+3jjs +15jjs+20j
=1
Thus
K =
js +1;j3jjs+1+j3jjs+15jjs+40j
js +3j
In this case we are not really sure that a break-in and break-out pointare
present. When we use MATLAB to drawtheroot locus at the end of the
problem we will nd out for sure.
To nd angle of departure, whichmayormaynot help us decide if
break-in and break-out points are present, werefer to Figure 6 and invoke
the angle condition
;
1
;
2
;
3
;
4
= ;180
:
As the circle shrinks in radius all the angles except
1
can be computed.
That is,
1
= tan
;1
(3=2)+ 180
;tan
;1
(3=14);tan
;1
(3=19);90
= 56:31
+ 180
;90
;12:09
;8:9726
= 125:24
The angle of departure is really a neutral indicator in this case. The
MATLAB dialogue
2
Im(s)
Re(s)
-20
Im(s)
Re(s)
-3-20
(a)
(b)
-15
-15
-3
Figure 3: TwoPotential Root Loci
3
Im(s)
Re(s)-3
Re(s)-3
K
break-out break-in
Im(s)
Re(s)-15
Re(s)
-15
K
(a)
(b)
X
X
-15
-15
-20
-20
-3
-3
X
X
X
X
-20
-20
Figure 4: Gain Versus Position for s 2 [;15;;;3]
4
Im(s)
Re(s)
-15
s
-20
s + 20
s + 25
s+3
s + 1 + j3
s +1 - j3
-3
Figure 5: Computation of Gain Along Real Axis
Im(s)
Re(s)
-15
3
-20
θ
1
θ
2
α
θ
3
θ
4
?3
?3
Figure 6: Calculation of angle of Departure
5
-25 -20 -15 -10 -5 0 5
-10
-8
-6
-4
-2
0
2
4
6
8
10
Real Axis
I
mag
A
x
i
s
Figure 7: MATLAB generated root locus
EDU>gh = zpk([-3],[-15 -20 -1+j*3 -1-j*3],10)
Zero/pole/gain:
10 (s+3)
-----------------------------
(s+15) (s+20) (s^2 + 2s + 10)
EDU>rlocus(gh)
EDU>print -deps rl5833f.eps
EDU>
draws the root locus shown in Figure 7. In this case, wesee that the poles
and GH are clusted tightly enough that there are no break-in and break-out
points on the interval [;15;;;3].
6