ECE 425: STATE SPACE DESIGN AND DIGITAL CONTROL
ASSIGNMENT IV (CAT 11)
Attempt all questions
Show all your workings
Due date: 28th March 2023
1. Consider the pulse-transfer-function systems below,
𝑧 −1 (1+𝑧 −1 )
a) 𝐺(𝑧) = (1+0.5𝑧−1 )(1+0.5𝑧−1 )
1+0.8𝑧 −1
b) 𝐺(𝑧) = 1−𝑧−1 +0.5𝑧−2
Obtain the state-space representation of the systems above in the following forms:
i. Controllable canonical form
ii. Observable canonical form
iii. Diagonal canonical form
2. Obtain a state-space representation of the following pulse-transfer-function such that the state matrix is a
diagonal matrix
𝑌(𝑠) 1
=
𝑈(𝑠) (𝑧 + 1)(𝑧 + 2)(𝑧 + 3)
Then obtain the initial state variables 𝑥1 (𝑘), 𝑥2 (𝑘), and 𝑥3 (𝑘) in terms of 𝑦(0), 𝑦(1), and 𝑦(2).
3. Consider a system defined by
𝑥1 (𝑘 + 1) 𝑏 𝑥1 (𝑘)
𝑎 1
[ ]=[ ][ ] + [ ] 𝑢(𝑘)
𝑥2 (𝑘 + 1) 𝑑 𝑥2 (𝑘)
𝑐 1
𝑥1 (𝑘)
𝑦(𝑘) = [1 0] [ ]
𝑥2 (𝑘)
Determine the conditions on 𝑎, 𝑏, 𝑐, and 𝑑 for complete state controllability and complete observability.
4. Consider a system defined by
𝑥 (𝑘 + 1) 0 1 𝑥1 (𝑘) 1
[ 1 ]=[ ][ ]+[ ] 𝑢(𝑘)
𝑥2 (𝑘 + 1) −0.16 −1 𝑥2 (𝑘) −0.5
𝑥 (0) 1
[ 1 ]=[ ]
𝑥2 (0) −1
If the system is completely state controllable. Determine a sequence of control signals 𝑢(0) and 𝑢(1) such that
the state 𝑥(2) becomes
𝑥 (2) −1
[ 1 ]=[ ]
𝑥2 (2) 2
5. Consider a system defined by
𝑥 (𝑘 + 1) 0 1 𝑥1 (𝑘) 1
[ 1 ]=[ ][ ] + [ ] 𝑢(𝑘)
𝑥2 (𝑘 + 1) −0.16 −1 𝑥2 (𝑘) 1
𝑥1 (𝑘)
𝑦(𝑘) = [1 0] [ ]
𝑥2 (𝑘)
If the outputs 𝑦(0) = 1, and 𝑦(1) = 2 are observed and the control signals are 𝑢(0) = 2, and 𝑢(1) = −1.
i. Determine the initial state 𝑥(0).
ii. Determine states 𝑥(1) and 𝑥(2).
6. Consider the system below
𝒙(𝑘 + 1) = 𝑮[𝒙(𝑘) + 𝑪∗ 𝒖(𝑘)]
𝑦(𝑘) = 𝑪𝒙(𝑘)
0100
Where, 𝑥(𝑘) = state vector (4-vector), 𝑢(𝑘) = control signal (scalar) 𝑦(𝑘) = output signal (scalar), 𝐺 = [0010], and
0001
1000
𝐶 = [1 0 0 0].
i. Show that the system is completely state controllable and completely observable.
ii. Show that at any initial state 𝑥(0) every state vector can be brought to the origin in at most four sampling
periods if and only if the control signal is given by 𝑢(𝑘) = −𝑪𝒙(𝑘).
7. Consider the double-integrator system
𝒙(𝑘 + 1) = 𝑮𝒙(𝑘) + 𝑯𝒖(𝑘)
𝑇2
1 𝑇
Such that 𝑮 = [ ],𝐻 = [2] and 𝑇 is the sampling period. Determine the state feedback gain matrix 𝑲, if the
0 1 𝑇
is desired closed-loop poles are located at 𝑧 = 𝜇1 and 𝑧 = 𝜇2 , and the state feedback control is given by
𝑢(𝑘𝑇) = −𝑲𝒙(𝑘𝑇).
8. Consider the system defined by
𝑥1 (𝑘 + 1) 0 1 0 𝑥1 (𝑘) 1
𝑥
[ 2 (𝑘 + 1) ]=[ 0 0 𝑥 (𝑘)
1] [ 2 ] + [1] 𝑢(𝑘)
𝑥3 (𝑘 + 1) −0.16 0.84 0 𝑥3 (𝑘) 1
If the control signal is 𝑢(𝑘) = −𝑲𝒙(𝑘) determine the state feedback gain matrix 𝑲 such that the closed-loop
system will exhibit the deadbeat response to any initial state 𝑥(0).
9. Consider the system defined by
𝒙(𝑘 + 1) = 𝑮𝒙(𝑘) + 𝑯𝒖(𝑘)
𝑦(𝑘) = 𝑪𝒙(𝑘)
Where, 𝑥(𝑘) = state vector (2-vector), 𝑢(𝑘) = control signal (scalar), 𝑦(𝑘) = output signal (scalar), 𝐺 =
0 1 0
[ ],𝐻 = [ ] and 𝐶 = [1 1]. Design a current observer for the system. It is desired that the response to
−0.16 −1 1
the initial observer error be deadbeat.
10. Consider the system defined by
𝒙(𝑘 + 1) = 𝑮𝒙(𝑘) + 𝑯𝒖(𝑘)
𝑦(𝑘) = 𝑪𝒙(𝑘)
Where, 𝑥(𝑘) = state vector (3-vector), 𝑢(𝑘) = control signal (scalar), 𝑦(𝑘) = output signal (scalar), 𝐺 =
0 0 −0.25 1
[1 0 0 ],𝐻 = [0] and 𝐶 = [1 0 0]. Assuming that the output 𝑦(𝑘) is measurable, design a minimum-order
0 1 0.5 1
observer such that the response to the initial observer error is deadbeat.
