5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.1. Introduction
1. We begin by introducing the concepts of controllability and
observability
2. Then discuss the pole placement design technique using the
powerful notion of state variable feedback.
3. Various methods for determining the state variable feedback
gain matrix, such as direct substitution, transformation
matrix and Ackermann's formula methods are discussed.
4. Similar methods are applicable to Observer design in cases
where some states are not available for feedback.
5. A state variable compensator can be realized by combining
the full state feedback law and the observer.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 1

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.2. Controllability and Observability

 Controllability tells us whether it is at all possible to

control all the states of the system completely by suitable
choice of input.
 Observability indicates whether it is possible to
reconstruct the states of a system from its input and
outputs.
 Conformity to controllability is necessary in the design of
control systems by pole placement technique.
 On the other hand, observability is a precondition to the
observer design.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 2

1
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.2.1 Controllability
 A system is completely controllable if there exists a control vector 𝑢(𝑡) that
can transfer any initial state 𝑥(t ) to any other desired location 𝑥(𝑡) over a
finite time interval, t ≤ 𝑡 ≤ 𝑇.
 Consider a linear time invariant system defined by

𝑥̇ 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑢 𝑡 (2.1a)

𝑦 𝑡 = 𝐶𝑥 𝑡 (2.1b)

 If the initial state is specified at time t , the solution of the state equation
yields
𝑥 𝑡 = ϕ t 𝑥 0 + ∫ ϕ t − 𝜏 𝐵𝑢 𝜏 𝑑𝜏 (2.2)

where ϕ t = 𝑒
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 3

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.2.1 Controllability
 For the system to be controllable

𝑥 0 =ϕ t 𝑥 0 + ϕ t − 𝜏 𝐵𝑢 𝜏 𝑑𝜏 = 0 (2.3)
with finite t

 For a LTI continuous time system described by eqn. (2.1), a sufficient condition for
complete state controllability is that the 𝑛 × 𝑛 controllability matrix 𝑀, defined by

𝑀= 𝐵 𝐴𝐵 𝐴 𝐵…..𝐴 𝐵 (2.4)

contains 𝑛 linearly independent row or column vectors, i.e. is of rank 𝑛 (that is,
nonsingular).
 If a system is not completely controllable, it implies that it has one or more natural
modes that cannot be affected by the input directly or indirectly.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 4

2
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.2.2 Observability
 The linear time invariant system is said to be observable if at time t , the system state 𝑥(t )
can be exactly determined from observation of the output 𝑦(𝑡) over a finite time interval
t ≤ 𝑡 ≤ 𝑇.

𝑦 𝑡 = 𝐶𝑥 𝑡 (2.5a)
= Cϕ t 𝑥 0 + C ∫ ϕ t − 𝜏 𝐵𝑢 𝜏 𝑑𝜏 (2.5b)

 Thus given 𝑢 𝑡 and 𝑦 𝑡 for t ≤ 𝑡 ≤ 𝑇 with 𝑇 being some finite value, the system is
observable if equation (2.5b) can be solved for 𝑥 0
 The system is observable if the observability matrix N is nonsingular i.e. the rank of N is
equal to 𝑛
𝐶
𝑁= 𝐶𝐴 (2.6)
⋮
𝐶𝐴
 The eqns. represented by (2.4) and (2.6) are referred to as Kalman’s tests for controllability
and observability.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 5

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

Example 2.1: Controllability and Observability of a two-state system
Determine whether the system represented by the given state space is
i. controllable
ii. observable
𝑥̇ 0.5 0 𝑥 0
= + 𝑢(𝑡)
𝑥̇ 0 −2 𝑥 1
𝑥
𝑦 𝑡 = 0 1 𝑥
Solution
The controllability matrix is given by

𝑀= 𝐵 𝐴𝐵

0.5 0 0 0
𝐴𝐵 = =
0 −2 1 −2

0 0 0 0
𝑀=→ 𝑀 = =0
1 −2 1 −2
The matrix is singular hence the system is uncontrollable.

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 6

3
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

Example 2.1: Controllability and Observability of a two-state system
Solution Cont’
 This is more obvious if we write the two differential equations separately as

𝑥̇ = 0.5𝑥

𝑥̇ = −2𝑥 + 𝑢(𝑡)

 It is evident that whereas 𝑥 can be changed by 𝑢(𝑡) the state 𝑥 is unaffected by our choice of the
inputs since it is not coupled either directly to the input or to the state 𝑥 hence this state of 𝑥 (0)𝑒 .
is uncontrollable.
 NB: On the other hand if we had

𝑥̇ = 0.5𝑥 + 𝑥
𝑥̇ = −2𝑥 + 𝑢(𝑡)
The controllability matrix is obtained as
0 1 0 1
𝑀= → 𝑀 = = −1
1 −2 1 −2

The matrix is nonsingular hence the system is controllable. 𝑥 can be controlled indirectly via 𝑥

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 7

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

Example 2.1: Controllability and Observability of a two-state system

Solution Cont’
(ii) The observability matrix is given by
𝐶
𝑁=
𝐶𝐴

0.5 0
𝐶𝐴 = 0 1 = 0 −2
0 −2
0 1
𝑁 = =0
0 −2

 The matrix is singular and therefore the system is unobservable.

 The state 𝑥 does not affect the output nor does it affect the state 𝑥 which
is coupled to the output.

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 8

4
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

Exercise: Controllability and Observability of a three-state system

(i) Show that the following system is not completely observable

(ii) Also, check the system for controllability.

𝑥̇ 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑢 𝑡

𝑦 𝑡 = 𝐶𝑥 𝑡
Where

𝑥 0 1 0 0
𝑥= 𝑥 ,A = 0 0 1 , 𝐵 = 0 ,𝐶 = 4 5 1
𝑥 −6 −11 −6 1

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 9

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.3. State Variable Design
The state variable design is based on the pole placement design technique
and involves three steps:
1. Step I: Assuming that all the state variables are measurable and available
for feedback, a full-state feedback control law can be obtained.
2. Step II: If some states are not directly sensed and available as outputs, an
observer is constructed to estimate such states. Observers can either be
full state observers or reduced-order observers.
3. Step III: This last step entails combining the observer with the full state
feedback control law. The state-variable controller (full-state control law
plus the observer) is usually referred to as a compensator, whose form is as
depicted in Figure 2.1.
Furthermore, it is possible to consider reference inputs to the state variable
compensator to complete the design.
 It also enables inclusion of initial conditions if necessary.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 10

5
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.3. State Variable Design

Fig.2.1 State Variable Controller Design

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 11

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.3.1 Pole Placement Design Technique
 Pole placement design technique (also known as State Variable feedback Design) is
based on the state model of the system.
 We assume that all the state variables are measurable and are available for feedback.
 Consider a system defined by the state model equations

𝑥̇ 𝑡 = 𝐴𝑥 𝑡 + 𝐵𝑢 𝑡 (2.7)

𝑦 𝑡 = 𝐶𝑥 𝑡

 Select a control law or control rule 𝑢(𝑡) which is made of a linear function of the states and
of the form:
𝐮 t = 𝐫 − 𝐊x t (2.8)

 In eqn. (2.8), 𝐫(t) is a vector of desired state variables and 𝐊 is known as the state variable
feedback gain matrix.
 The state variable block diagram representation of equations (2.7) and (2.8) is as given in
Fig.2.2.
EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 12

6
 5/19/2024

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.3.1 Pole Placement Design Technique
 x
 Substituting equation (2.8) into (2.7), we get
r  u 𝑥̇ y
B C

∫
ẋ t = Ax t + B 𝐫 − 𝐊x t 

= 𝐀 − 𝐁𝐊 x t + B𝐫 (2.9)

where the matrix 𝐀 − 𝐁𝐊 is the closed-loop A

system matrix.
 The characteristic equation for the system of
eqn. (2.7) is given by
(s𝐈 − 𝐀) = 0 (2.10)
K
 The roots of eqn (2.10) are the open-loop poles Fig.2.2 Control using State Variable Feedback
or eigenvalues.
 For the closed-loop system defined by eqn (2.9), the characteristic equation is
s𝐈 − 𝐀 + 𝐁𝐊 = 0 (2.11)

whose roots are the closed-loop poles or eigenvalues.

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 13

2. DESIGN OF CONTROL SYSTEMS IN STATE SPACE

2.3.1 Pole Placement Design Technique
Regulator design by pole placement
 In the pole placement design problem, the objective is to determine a value of the control law
𝐊 that will allow the poles of the closed loop system to be placed in any desirable location.
 Noting that with a regulator, 𝐫 t = 𝟎, from eqn (2.9), the control law becomes

𝑢 𝑡 = −𝑘𝑥 𝑡 (2.12𝑎)

Or
𝑢 𝑡 = −𝑘 𝑥 𝑡 − 𝑘 𝑥 𝑡 − … − 𝑘 𝑥 𝑡 (2.12𝑏)

 We see that the control 𝑢 𝑡 will drive the system from a set of initial conditions 𝑥 0 to a set of
zero states at time 𝑡 , i.e. 𝑥 𝑡 = 0.
 The design problem is the specification of the desired root locations of the system’s
characteristic equation and the calculation of the gains 𝑘 to yield these desired root locations.
 A necessary and sufficient condition that the closed-loop poles can be placed at any arbitrary
location in the s-plane is that the system must be completely state controllable.

EEE 2502 Control Engineering III Lecture Notes by Dr. Linus. A. Aloo
5/19/2024 14