NPTEL Online Course (July-Nov 2017)

Control Engineering

Assignment 5 Solution

Q.1
Let s be a point near the pole on +ve imaginary axis. Let θ be the angle of departure.
Then,

θ = 180o + (−90o + 90o ) − (90o + 90o ),

⇒ θ = 0o

∴ The angle of departure is 0o which satisfies the root locus plot shown in Figure 1.

Figure 1: Valid root locus for the given poles and zeros

Q.2
Given,

K(s + 23
G(s) = , H(s) = 1
s2 (s + 2)

Step 1
The loop transfer function G(s)H(s) has three poles at s = 0, s = 0, s = −2. Therefore, the root locus have three
branches.

Step 2
The three branches originate from the poles, p1 = 0, p2 = 0, p3 = −2 for K = 0 and terminate at the zero z1 = − 23 and
infinity.

Step 3
All the points on the real axis between − 23 and −2 lie on the root locus since the number of poles and zeroes to the right
is 3 (an odd number). Therefore, the branch originating from p3 = −2 terminates at z1 = − 32 .
Step 4
The two branches that terminate at infinity do so along the asymptote with the angles
(2q + 1)180o
φA = q = 0, 1, ..., (n − m − 1)
n−m
(2q + 1)180o
= q = 0, 1
2

For q = 0, φA = 90o
For q = 1, φA = 270o

Step 5
The asymptotes meet at the centroid, σA
Sum of real parts of poles − Sum of real parts of zeros
σA =
Number of poles − Number of zeros
(0 + 0 − 2) − (− 23 )
=
3−1
2
=−
3

Step 6
dK
The break-away point is determined from the equation ds =0
Here,
s2 (s + 2)
K =−
s + 23
dK (3s2 + 4s)(s + 23 ) − (s3 + 2s2 )1
∴ =
ds (s + 32 )2
s (3s2 + 6s + 83 ) − (s2 + 2s)


=
2
s + 23
2s 3s2 + 6s + 4


=
2
3 s + 23
dK
For ds = 0,

2s(3s2 + 6s + 4) = 0
√
2
⇒ s = 0, −1 ± j
3
Only the point s = 0 lie on the root locus. Hence, the break-away point is s = 0.

Step 7
The angle of departure for the poles p2 and p3 are
φ p2,3 = 180o + 0o = 180o
The root locus of the system is plotted below in Figure 2

Q.3
The given pole-zero plot is shown in Figure 3.
Poles are at s = −2, −3 and zeros are at s = 1 ± j.
Then the transfer function can be written as



K s − (1 + j) s − (1 − j) K(s2 − 2s + 2)
T (s) = =
(s + 2)(s + 3) (s + 2)(s + 3)
 Figure 2: Root locus of the system in Q.2

Q.4
From Q.3 we get

s2 + 5s + 6
K=−
s2 − 2s + 2
Now, for break-away point or break-in point

dK −7s2 − 8s + 22
= =0
ds (s2 − 2s + 2)2
⇒ (7s2 + 8s − 22) = 0
⇒ s = 1.29, −2.43

The point s = 1.29 does not lie on the root-locus. Hence, it is not a break point.
Now,

2 −7x3 − 12x2 + 66x − 36

d2K
=−
ds2 (x2 − 2x + 2)3
d 2 K 

and, = −0.16 < 0
ds2 s=−2.43

i.e. at s = −2.43, K attains maximum value. Therefore, s = −2.43 is a break-away point.

Alternative Method
From the given options s = −1.29, −2.43. All points on the real axis between −3 and −2 is a part of root locus since
number of poles and zeros to the right is 3 (an odd number). Hence, only −2.43 is a part of the root locus.
Now, the system has poles at −3, −2 and zeros at −1 ± j i.e the branches originate from the poles at −2, −3 and terminate
at −1 ± j. Definitely, the branches meet at a point between −2 and −3 and break-away to the zeros.
∴ −2.43 is a break-away point.
 Figure 3: Pole-Zero plot of Q.3

Q.5
Option (c) is correct.
Satisfy all property. Root locus start from pole and end on zeros. On real axis locus exist to the left of odd number of
poles and zeros. Locus has symmetry.

Q.6
Break-away and break-in method can also be determined using the transition method [1], illustrated in Appendix A.1.
Then,
1 1 1 1
+ = +
(σb + 1) (σb + 2) (σb + 5) (σb + 6)
⇒ 2σb 2 + 14σb + 17 = 0
⇒ σb = −1.563, −5.437

Since, the branches originate from the poles situated at −5 and −6 the break-away point is −5.437. While, the branches
terminate at the zeros situated at −2 and −1, the break-in point is −1.563.

Q.7
The open loop gain K at any point s0 on the root locus is determined as

∏ni=1 |s0 + pi |
K=
∏mi=1 |s0 + zi |

Here, poles are p1 = 0, p2 = −4, p3 = −12, p4 = −20 and zeros are z1 = 8.

10 × 6 × 2 × 10
∴ K= = 600
2

Q.8
The given system is a positive feedback system. The odd multiplying factor (2q + 1) in the angle condition of Evans’
is replaced by even multiplying factor of 2q, where q = 0, 1 2, · · · , (n − m − 1). All points on the real axis having odd
number of poles and zeros to it’s left is a part of the root locus. Therefore, the branch originating from the pole s = −3
terminates at the zero located at s = −2 and the branch originating from the pole s = −1 terminates at +∞. Hence, the
root locus of the given positive feedback system is as shown in Figure 4.
 Figure 4: Root locus of the system in Q.8

Q.9
Applying the transition method for determining break-away/in points we get,
1 1 1
= +
σb + 6 σb + 3 σb + 5
⇒ σb 2 + 8σb + 15 = 2σb 2 + 20σb + 48
⇒ σb 2 + 12σb + 33 = 0
⇒ σb = −4.267, −7.732 (correct upto three decimal places)

The breakaway point lies between the poles and the break-in point lies between the zeros. Hence, the break-away point is
−4.267 while the break-in point is −7.732.

Q.10
Here,

K = −s(s + 4)(s2 + 4s + 5)
dK h i
∴ = − (2s + 4)(s2 + 4s + 5) + (s2 + 4s)(2s + 4) = 0
ds
⇒ (s + 2)(2s2 + 8s + 5) = 0
⇒ s = −0.775, −2, −3.225

So the break points are −0.775, −2, −3.225.

Now,
d2K
= − (2s2 + 8s + 5) + (s + 2)(4s + 8) = −(6s2 + 24s + 21)
 
ds 2

d 2 K 
∴ 2 ≈ −6 < 0
ds s=−0.775
d 2 K 

≈3>0
ds2 s=−2
d 2 K 

≈ −6 < 0
ds2 s=−3.225
Hence, −0.775 and −3.225 are the breakaway points and −2 is the break-in point i.e. 2 breakaway point and 1 break-in
point.

Q.11
The given characteristic equation can be expressed as

K(s2 + 4)
1+ =0
(s + 2)2 (s + 5)(s + 6)
Poles are at s = −2, −2, −5, −6 and zeros are at s = − j2, j2.
Intersection of the asymptotes is

(−2 − 2 − 5 − 6) − (0)
σA = = −7.5
4−2
i.e. (−7.5, 0).

Q.12
All points on the real axis, lying between −2 and −4 are part of the root locus. Also, entire real axis to the left of −6 is a
part of the root locus. One branch terminate at −6 and the other at infinity. Now, the angle of asymptote is given by

φA = 180o

and the centroid of the asymptotes is

(−4 − 2) − (−6)
σA = =0
2−1
Hence, the real axis is the asymptote and one branch moves toward −∞ along the asymptote.
Now, break points can be determined from
1 1 1
= +
σb + 6 σb + 2 σb + 4
σb + 6σb + 8 = 2σb 2 + 18σb + 36
2

σb 2 + 12σb + 28 = 0
σb ≈ −3.17, −8.83

Both the point lie on the root locus, and thus are root break points. So the root locus intersect the real axis twice. Hence,
statement (I) and (III) are correct.

Q.13
The characteristic equation can also be expressed as

s(s + 1)(s + 2) + K = 0
K
⇒ 1+ =0
s(s + 1)(s + 2)
Therefore the centroid of the asymptotes is

(0 − 1 − 2) − (0)
σA = = −1
3−0

Q.14
The open loop transfer function of the unity feedback system is given by

K(s + 3)
G(s) =
s(s + 1)(s + 2)(s + 4)
Then the loop transfer function is

G(s)H(s) = G(s)

Then poles of the loop transfer function are 0, −1, −2, −4 and one zero at −3. The angle of asymptotes is determined as

(2q + 1)180o
φA = , q = 0, 1, · · · , (n − m − 1)
n−m
(2q + 1)180o
= , q = 0, 1, 2
3
= 60o , 180o , 300o
Q.15
Given test points are s1 = −2 + j3 and s2 = −2 + j √12 .
For s1 , angle contribution of the system poles and zeros on the test point is
3 π
arctan( ) + arctan(3) − − π − arctan(3) ≈ −70.56o


2 2
For s2 , angle contribution of the system poles and zeros on the test point is
1 1 π  1 
arctan( √ ) + arctan( √ ) − − π − arctan( √ ) = −180o
2 2 2 2 2
The angle contribution of all system poles and zeros on a point on the root locus is always an odd multiple of ±180o .
From the angle contributions computed above it can be concluded that s2 is a point on the root locus but not s1 .

Q.16
Given loop transfer function is

K(s + 2)
G(s)H(s) =
(s + 3)(s3 + 2s + 2)
Poles of the system are p1 = −3, p2 = −1 + j, p3 − 1 − j and zero is z1 = −2. Angle contribution of the poles and zeros
on p2 is given by
1
φ = θz1 p2 − θ p1 p2 − θ p3 p2 = arctan(1) − arctan( ) − 90o ≈ −71.565o
2
o o
 The angle of departure for the pole p2 is 180 + φ = 108.435 . Similarly, the angle of departure for the pole p3 is
∴
180o + arctan(−1) − arctan(− 21 ) − (−90o ) ≈ 251.565o = −108.435o

Q.17
Given that no part of the root locus lies in the RHP. Then both a and b are negative.There are two poles, p1 = 0 and
p2 = −b, and one zero, z1 = a. If p2 = z1 then the terms (s + a) and (s + b) cancel out and the transfer function resembles
an integrator whose root locus is nothing but a straight line originating at the pole located at the origin and moving towards
infinity along the real axis. So, for a circle to exist as a part of the root locus there shouldn’t be any equality condition
between a and b.
The root loci of the transfer function have two branches, originating at p1 and p2 and terminating at z1 and infinity.
The points on the real axis which is a part of the root locus depend on the values of a and b.
(2q+1)180o
The angle of asymptote is σA = 2−1 , where q = 0 i.e. σA = 180o .
dK
Break points are determined from ds = 0. Analysing the condition, it can be found that

dK
=0
ds
s2 + bs
 
d
⇒ − =0
ds s+a
(2s + b)(s + a) − (s2 + bs)1
⇒ =0
(s + a)2
⇒ s2 + 2as + ab = 0
p
⇒ s = −a ± a2 − ab

break points will exist if a2 − ab > 0 i.e. |a| > |b|. If a root locus of a transfer function consists a circle then it is

necessary that break points exists for the system. If not then there will be no circle. Consider the case when |a| < |b|.
Then z1 lie between p1 and p2 . Hence, all points on the real axis lying between z1 and p1 is a part of the root locus. Also,
all points on the real axis to the left of p2 is a part of root locus. Therefore, the branch originating from p1 terminates
at p1 , moving along the real axis while the branch originating from p2 terminates at −∞. Both the branches are straight
lines along the real axis. So, the necessary condition for the existence of a circle is |a| > |b|.
But is the condition also sufficient?
√
Let us consider |a| > |b|√and hence break points exist. 2

In that case, the breakaway point is (−a + a − ab) and the
2
break-in point is (−a − a − ab) Verify Yourself . Thus the branches originating from p1 and p2 meet and breaks
√ √
away at s = −a + a2 − ab and again meet or break-in at s = −a − a2 − ab.
Now, any point in the s-plane represented by s = σ + jω satisfies the Evan’s angle condition

∠G(s) = ±(2q + 1)180o , q = 0, 1, · · · , (n − m − 1)

where symbols have their usual meanings. Then,

(s + a)
∠ = ±180o
s(s + b)
⇒ ∠(σ + a + jω) − ∠(σ + jω) − ∠(σ + b + jω) = ±180o
 ω  ω   ω 
⇒ arctan − arctan − arctan = ±180o
σ +a σ σ +b
σ 2 + 2aσ + ω 2 + ab
 
⇒ arctan − 3 = ±180o
σ + (a + b)σ 2 + (ω 2 + ab)σ + (b2 − a2 )ω 2
⇒ σ 2 + 2aσ + ω 2 + ab = 0
⇒ (σ + a)2 + ω 2 = (a2 − ab)
p
⇒ (σ + a)2 + ω 2 = r2 , where r = a2 − ab.
√
This is an equation of a circle with centre (−a, 0) and radius a2 − ab. Therefore the condition |a| > |b| is both necessary
and sufficient.

Q.18
Given characteristic equation is

s(s + 1)(s + 3) + K(s + 2) = 0

s3 + 4s2 + (K + 3)s + 2K = 0

Routh table for the characteristic equation, s3 + 4s2 + (K + 3)s + 2K = 0

s3 1 K +3
s2 4 2K
K+6
s1 2
s0 2K
For K > 0, the system is stable i.e. there are no roots in the RHP or no repeated roots on the imaginary axis. Therefore,
option (d) is incorrect. Thus the characteristic equation can be equivalently expressed as

K(s + 2)
1+ =0
s(s + 1)(s + 3)

Hence, the loop transfer function is given by

K(s + 2)
G(s)H(s) =
s(s + 1)(s + 3)

Step 1
Poles of the loop transfer function are p1 = 0, p2 = −1, p3 = −3 and zero is at z1 = −2. Therefore, the root locus have 3
branches

Step 2
The three branches originate from the poles p1 = 0, p2 = −1, p3 = −3. One branch terminates at the zero z1 = −2 and
the other two branches terminate at infinity.
Step 3
All the points on the real axis lying between 0 and −1 and between −2 and −3 are a part of the root locus since the
number of poles and zeros to the right of these points is odd.

Step 4
The angle of asymptotes is given by

(2q + 1)180o
φA = , q = 0, 1
3−1
= 90o , 270o

Step 5
The centroid of the asymptotes is given by

(0 − 1 − 3) − (−2)
σA = = −1
2
Therefore, option (c) is correct.

Step 6
Applying the transition method for determining break points,
1 1 1 1
= + +
σb + 2 σb σb + 1 σb + 3
⇒ σb 3 + 4σb 2 + 3σb = (σb + 2) (σb 2 + 4σb + 3) + (σb 2 + 3σb ) + (σb 2 + σb )
 

⇒ σb 3 + 4σb 2 + 3σb = (σb + 2)(3σb 2 + 8σb + 3)

⇒ σb 3 + 4σb 2 + 3σb = 3σb 3 + 14σb 2 + 19σb + 6
⇒ 2σb 3 + 10σb 2 + 16σb + 6 = 0
⇒ σb = − 0.534, −2.233 ± j0.793

Then the break points is −0.534 since the other two point doesn’t lie on the root locus (Verify yourself). Therefore option
(b) is wrong.
Now, the root locus of the system is plotted based on the information gathered from the six steps as shown in Figure 5
It can be observed that the roots can be complex also. Therefore, option (a) is wrong.

Q.19
The closed loop transfer function is given by
1
1 2
T (s) = 2
= s +1αs
s + αs + 1 1 + s2 +1

1
Thus an equivalent system has G(s) = s2 +1
and H(s) = αs. Therefore, the loop transfer function G(s)H(s) is given by

αs
G(s)H(s) =
s2 + 1
The root locus is plotted with α as the parameter. Then the poles are at ± j and zero at s = 0. There are two branches in
the root locus originating from the poles at ± j and terminating at 0 and ∞ respectively. The entire real axis in the LHP is
a part of the root locus. The angle of asymptote will be 180o and centroid at 0. The breakaway point lies at

dα (s2 + 1) − 2s(s) s2 − 1
=− = =0
ds (s2 + 1)2 (s2 + 1)2

Therefore, the break-away point is σb = −1. The angle of departure of the pole at j is 180o + 90o − 90o = 180o and that
of the pole at − j is 180o + 90o − 90o = 180o . The root locus plot is shown in Figure 6
 Figure 5: Root locus plot of Q.18

Q.20
Characteristic equation of the system is given by

s(s + 2)(s + 10) + 2(s + α) = 0

⇒ s3 + 12s2 + 22s + 2α = 0
2 α
An equivalent system for the characteristic equation is obtained for G(s) = s2 +12s+22
and H(s) = s i.e. the loop transfer
function is given by
2α
G(s)H(s) =
s(s2 + 12s + 22)

Now, the root locus of the loop transfer function can be plotted by varying α. The poles of the transfer function are at
s = 0, −2.258, −9.742. Then the centroid of the asymptotes is

(0 − 2.258 − 9.742) − (0)

σA = = −4
3

References
[1] Gene F Franklin, J David Powell, Abbas Emami-Naeini, and J David Powell. Feedback control of dynamic systems,
volume 3. Addison-Wesley Reading, MA, 1994.

A Appendix
A.1 Transition Method of Determining Break-away/in points
The transition method for finding real-breakaway and break-in points without differentiating can be found in [1]. The
method is derived showing that the natural logarithm of
1
has a zero derivative at the same value of σb as
G σb H σb
 Figure 6: Root locus plot of Q.19

1
.
G σb H σb

Consider,
   
d 1

d 1
ln

= G σb H σb

=0 (1)
dσb G σb H σb dσb G σb H σb



Since G σb H σb is not zero at the breakaway or break-in points, then
 
d 1


=0 (2)
dσb G σb H σb

Hence, the same value of σb is obtained for both

   
d 1 d 1
ln

=0 and

=0
dσb G σb H σb dσb G σb H σb

Now, it is known that break-away or break-in points are derived from dK

ds = 0 which can also be written as
 
dK d −1
= =0 (3)
ds ds G(s)H(s)
 
d 1
⇒ − =0 (4)
ds G(s)H(s)
 
d 1
⇒ =0 (5)
ds G(s)H(s)
∴ Solving equation 1 will give break-away and break-in points. Now, equation 1 can be expressed as:
   
d 1 d (σb + p1 )(σb + p2 ) · · · (σb + pn )
ln

= ln
dσb G σb H σb dσb (σb + z1 )(σb + z2 ) · · · (σb + zm )
d 
= ln(σb + p1 ) + ln(σb + p2 ) + · · · + ln(σb + pn )
dσb

− ln(σb + z1 ) − ln(σb + z2 ) − · · · − ln(σb + zn )
   
1 1 1 1 1 1
= + +···+ − − −···−
σb + p1 σb + p2 σb + pn σb + z1 σb + z2 σb + zm
=0 (6)

or,
n   m  
1 1
∑ = ∑ (7)
i=1 σb + pi i=1 σb + zi

∴ Solving equation 7 leads to the break-away point and break-in point. To check which one is break-away and which one
is break-in point we have to find out whether K is maximum or minimum at those points, respectively.