EE207/23/Min2

EE207: Control Engineering

Minor Test II

Time: 2 hrs (Open Book) Marks: 20

____________________________________________________________________________________

1 The known OLTF for a system is similar to what was done in our lectures as the first example on root loci:

G OL (s ) = 1
s(s + 10 )
The unity negative feedback CLTF includes a cascaded PI controller with the usual parameters KP and TI,
which are maintained such that all CL poles have identical real components. For the parameters so set,
i. Obtain an expression for the real component that applies to all CL poles.
ii. Obtain an expression for natural frequency of oscillatory components in the CLTF.
iii. Obtain the required relation between KP and TI. ...(2 marks for each part)

2 A digital system (that has no analog origin), has its OLTF as

G OL (z ) = K
(z + 0.9 )(z − 0.9 )
In order to avoid aliasing and ringing without compromise on stability levels, it is decided to shift the poles for
unity negative feedback CLTF to !j0.9 rad/s.
i. In case the change is possible, find the value of K that can achieve the specification.
ii. In terms of the sampling time T, what are the natural frequencies of the OLTF ?
iii. In terms of the sampling time T, what are the natural frequencies of the CLTF ?
iv. If K is increased beyond that required in “i”, at what value will the system become marginally stable
(critical gain in digital systems terminology) ?
...(2 marks for each part)

3 A digital second order linear SISO system has matrices F, G, and C given by
2 5 0
F= ; G= ; C= 3 0
1 3 9
With the state space defined for state vector x(k), consider transformation to state space for state vector z(k)
such that
z 1 (k ) å x 1 (k ) + x 2 (k )
z 2 (k ) å x 1 (k ) − x 2 (k )
Reason out if the transformation is possible or not, explain why so, with proper analytical reasoning and
justification.

In case the transformation is possible, obtain the transformed matrices F*, G*, and C*, for the system in terms of
z(k).
...(6)
 EE207/23/Min2
EE207: Control Engineering

Solutions to Minor Test II

1
i. The characteristic equation for the CLTF is obtained as
T I s 2 (s + 10 ) + K P (1 + T I s ) = 0
u s 3 + 10s 2 + K P s + K P /T I = 0
Thus the CLTF must have one real pole and a complex conjugate pole-pair, giving the above a form:
(s + " )(s 2 + 2* n s + * 2n ) = 0
u s 3 + (" + 2* n )s 2 + (2* n " + * 2n )s + "* 2n = 0
Comparing the two,
" + 2* n = 10
2* n " + * 2n = K P
"* 2n = K P /T I
From the first relation, the real component of each pole must be 10/3.
ii. From the solution of “i”, and the second relation, the natural frequency of oscillations is obtained as
2 % 10 % 10 + * 2n = K P
3 3
u * n = K P − 200/9
iii. By the third relation, and the solutions to “i-ii”,
10 K − 200 = K P
P
3 9 TI
which is the required relation between KP and TI for the specified condition.
(2+2+2 marks)

2 The CLTF has its characteristic equation as

z 2 − 0.81 + K = 0
i. For the CLTF poles to be at !j0.9 rad/s, the characteristic equation should be
(z + j0.9 )(z − j0.9 ) = 0
u z 2 + 0.81 = 0
In order to obtain this condition, K = 2 x 0.81 = 1.62
ii. 0 rad/s (that is, DC), and !π/T rad/s .
iii. !π/(2T) rad/s.
iv. At the point of marginal stability, the characteristic equation of CLTF must become
(z + j1 )(z − j1 ) = 0
u z2 + 1 = 0
Comparing with the CL characteristic equation in terms of K, we obtainthe value to be K = 1.81.
(2+2+2+2 marks)

3. The transformation will be possible if both the transformation matrix and its inverse exist. By the proposed
relations between the state variables, we have
 EE207/23/Min2
1 1
z(k ) = $x(k )
1 −1
Inverse of the transformation matrix is obtained as
−1
1 1 1 −1 −1
=
1 −1 −1 − 1 −1 1

0.5 0.5
=
0.5 −0.5
The original state and output equations can be transformed as
2 5 0
x(k + 1 ) = $ x(k ) + $ u(k ) ; y(k ) = 3 0 $ x(k )
1 3 9

0.5 0.5 2 5 0.5 0.5 0

u $ z(k + 1 ) = $ $ z(k ) + $ u(k ) ;
0.5 −0.5 1 3 0.5 −0.5 9

0.5 0.5
y(k ) = 3 0 $ $ z(k )
0.5 −0.5

1 1 2 5 0.5 0.5 1 1 0
u z(k + 1 ) = $ $ $ z(k ) + $ $ u(k ) ;
1 −1 1 3 0.5 −0.5 1 −1 9

0.5 0.5
y(k ) = 3 0 $ $ z(k )
0.5 −0.5
Accordingly,
1 1 2 5 0.5 0.5 3 8 0.5 0.5 5.5 −2.5
F& = $ $ = $ =
1 −1 1 3 0.5 −0.5 1 2 0.5 −0.5 1.5 −0.5

1 1 0 9
G& = $ =
1 −1 9 −9

0.5 0.5
C& = 3 0 $ = 1.5 1.5
0.5 −0.5
(6 marks)