Part A – Designing the digital controller Gc(z) using indirect design method

Qn 1) Finding Damping Ratio, Natural Frequency and Characteristic Equation

−πζ

Overshoot M p=e √ 1−ζ

2

−πζ
√1−ζ 2
0.05=e
−πζ
ln 0.05=
√1−ζ 2
−2.996 ( √1−ζ )=−πζ
2

−2.996 (1 −ζ )=−πζ
1
2

−2.996+ 2.996 ζ=−πζ

πζ +2.996 ζ =2.996
2.996
ζ=
( π +2.996 )
2.996
ζ=
6.138
Damping Ratio ζ=0. 488

Settling Time ( 2 % ) T s =30 ms

4
T s=
ζ ωn
−3 4
30 x 10 =
( 0.488 ) ω n
4
0.488 ω n= −3
30 x 10
4
0.03
ω n=
0.488
133.333
ω n=
0.488
Natural Frequency ω n=273.22

Characteristic Equation:
2 2
s +2 ζ ω n s+ω n =0
2 2
s +2(0.488)(273.22) s+ 273.22 =0
2
s +266.66 s+74649.17=0
Qn 2) Finding compensator parameters in s-domain
Km
Tm
G p (s)=
(
s s+
1
Tm )
1
Let zero=
Tm
zero=50

70
G p (s)=
s (s+50)

Gc (s)=k c ( s+s+ zero

pole )

Let “ zero ”be a .

Let “ pole ” be b .

G c (s)=k c ( s+s+ ab )
OLTF=k c ( s+s+ ab ) x s(s+7050)
k c ( 70 ) ( s+ a )
OLTF=
s (s+50)( s+b)

Since a=50 ,

k c ( 70 ) ( s +50 )
OLTF=
s (s+50)( s+b)

k c ( 70 )
OLTF=
s (s+b)

N
CLTF=
N +D

70 k c
CLTF=
s ( s +b ) +70 k

70 k
CLTF= 2
s +sb +70 k
Closed Loop Characteristic Equation :
2 2
s +2 ζ ω n s+ω n =0
2
s + sb+70 k c =0

Deducing from sb=2ζ ω n s

b=2(0.488)(273.22)
b=266.66

Deducing from 70 k c =ωn2

2
70 k c =273.22
k c =1066.42

Therefore b=266.66 , a=50 , k c =1066.42

Compensator transfer function Gc ( s ) =k c ( ss +a+b )
G c (s)=1066.42 ( s +266.66
s+50
)
Qn 3) Finding the digital compensator using Tustin’s bilinear transformation

T s=0.002

2 z−1
s=
T s z +1

z−1
s=1000
z +1

Substituting s=1000
z−1
z +1
into Gc (s)=1066.42 (
s+50
s+266.66 )

( ( z+ 1 )

)
z−1
1000 +50
( z−1 )
G ( z ) =1066.42
c x
1000 ( )
z −1 ( z−1 )
+266.66
z+1

Gc ( z ) =1066.42
( 1000 ( z +1 ) +50( z −1)
1000 ( z +1 ) +266.66( z−1) )
G c ( z ) =1066.42 ( 10001000
z +1000+266.66 z−266.66 )
z +1000+50 z−50

Gc ( z ) =1066.42 ( 1266.66 z +733.34 )

1050 z +950

Gc ( z ) =1066.42
( 1050 (z+ 0.905)
1266.66 (z +0.579) )
Gc ( z ) =884.01 ( z+z+ 0.905
0.579 )

Finally ,∈the z−domain , a=0.905 , b=0.579 , k c =884.01

Qn 4) Simulation using Simulink and Matlab

Command in Matlab

Block diagram in Simulink

Simulation result to steady state in Simulink

Steady state value = 1, steady state error = 0

Settling time (2%) = 0.042sec

Overshoot = 1.319
 Part B – Designing the digital controller Gc(z) using direct design method

Qn 1) As found in Part A, question 1,

Natural Frequency ω n=273.22

Damping Ratio ζ=0. 488

Using the Natural Frequency and Damping Ratio, we can now find the desired closed-
loop poles in the s-domain,
s=−ζ ωn ± j ωn √ 1−ζ 2
s=−0.488(273.22)± j(273.22) √ 1−(−0.488)2
s=−133.33 ± j304.02
Desired poles S1 = -133.33 + j304.02 and S2 = -133.33 – j304.02.

From the s-domain to the z-domain,

sTs
z=e
(−ζω ± j ω √ 1−ζ ) T
, where T s = 2 ms
2

z=e n n s

0.002 (−133.33 ) ± j 0.002(304.02)

z=e e
( −0.267 ) (± j 0.608)
z=e e
z=¿ z∨( cos ∠ z ± jsin ∠ z )
z=¿ 0.765∨( cos ∠ 0.608 ± jsin ∠ 0.608 )

Desired poles in the Z-domain are

Z = 0.765 ∠ 0.608º = 0.765 ± j0.00812
Qn 2) Finding the plant, GHP(z).
G HP ( z )=(1−Z ) Z
−1G p (s)
s { }
−1
G HP ( z )=(1−Z ) Z 2
70
s ( s+ 50){ }
{ }
−1
(1−Z ) 1
G HP ( z )= Z 2
70 s (s +50)
(1−Z −1)
{ }
2
50
G HP ( z )= Z 2
3500 s (s +50)

a=50
aT =0.1
−aT
e =0.9048
−aT
aT e =0.0905

[ ]
−1
(1−Z ) ( 0.1−1+ 0.9048 ) z +(1−0.9048−0.0905)
G HP ( z )=
3500 ( 1−Z −1) ( z−1 ) (z −0.9048)

[ ]
−1
(1−Z ) 0.1 z−1 z+0.9048 z +1−0.9048−0.0905
G HP ( z )=
3500 ( 1−Z −1 ) ( z−1 ) (z−0.9048)

G HP ( z )=
1
[0.0048 z +0.0047
3500 ( z−1 )( z−0.9048) ]
G HP ( z )=
1
[ 0.0048(z +0.98)
3500 ( z−1 )( z−0.9048) ]
G HP ( z )=0.0014
[ z+ 0.98
( z−1 ) (z−0.9048) ]
Now that we have found G HP ( z ), we can derive our OLTF.
OLTF=0.0014 K c
[ z +0.98
( z−1 ) (z−b) ]
jm

Desired pole,
Z = 0.765 + j0.00812

∠ Pole 1
∠ Pole β
∠Zero
Re
Zero = -0.98 0 Pole β Pole = 1

∠ Zero=tan−1 ( 0.765+0.98
0.00812
)
∠ Zero=0.27 °

180−∠ Pole 1=tan−1 ( 1−0.765

0.00812
)
∠ Pole 1=178.02°

Conditions of angles,
Total angles of zeros – Total angles of poles = ±180°

Hence,
∠ Pole β=180 °−0.27 °−178.02 °
∠ Pole β=1.71 °

0.00812
tan (1.71 )=
0.765−Pole β
0.765−Pole β=0.272
Pole β=0.493
Now we can the compensator gain K c ,
0.0014 K c = [
( z −1 ) ( z−0.493)
z+ 0.98 ]
0.0014 K c = [
[ ( 0.765+ j 0.00812 )−1 ] [(1∠178.02)−0.493 ]
[(0.493 ∠ 1.71)+ 0.98] ]
0.0014 K c =¿ 0.238− j 0.016∨¿
K c =170

Lastly, we can now derive GC ( z ).

GC ( z )=170 ( z+z−0.493
0.9048
)

Command in Matlab
Block diagram in Simulink

Simulation result in Simulink