See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/351812789

Observer Design For State Variable Feedback Controller By Matlab

Conference Paper · May 2021

CITATIONS READS

0 56

2 authors, including:

Abdussalam Ali Ahmed

Bani Waleed University, Libya
58 PUBLICATIONS   50 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Using of Neural Network Controller to Improve Electric Vehicle Stability View project

Quality managment View project

All content following this page was uploaded by Abdussalam Ali Ahmed on 24 May 2021.

The user has requested enhancement of the downloaded file.

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

OBSERVER DESIGN FOR STATE VARIABLE FEEDBACK CONTROLLER BY MATLAB

Amhimmid Q. ALMABROUK
Mechatronics Department, Higher Institute of Engineering Technology, Bani Walid, Libya ORCİD: orcid.org/0000-0003-
2325-7071

Abdussalam Ali Ahmed

Mechanical Engineering Department, Bani Waleed University, Bani Waleed, Libya ORCİD: orcid.org/0000-0002-9221-2902

ABSTRACT
The drawbacks of frequency domain methods of design, using either root locus or frequency response
techniques, is that after designing the location of dominant second-order pair of poles, we keep our
fingers crossed, hoping that the higher-order poles do not affect the second-order approximation. What
we would like to be able to do is specify all closed-loop poles of the higher-order system. Frequency
domain methods of design do not allow us to specify all poles in systems of order higher than tow
because they do not allow for a sufficient number of unknow parameters to place all the closed-loop
poles uniquely. One gain to adjust, or compensator pole and zero to select, does not yield a sufficient
number of place all the closed-loop poles at desired locations. Remembers, to place n unknown
quantities, you need n adjustable parameters. State-space methods solve this problem by introducing
into the system (1) other adjustable parameters and (2) the technique for finding these parameter
values, so that we can property place all poles of the closed-loop system.
On the other hand, state-space methods do not allow the specification of close-loop zero location,
frequency domain methods do allow through placement of the lead compensator zero. This is a
disadvantage of state-space methods, since the location of the zero does affect the transient response.
Also, a state-space design may prove to be very sensitive to parameter changes.
Finally, there is a wide range of computational support for state-space methods; many software
packages support the matrix algebra required by the design process. However, as mentioned before,
the advantage of computer support are balanced by the loss of graphic insight into a design problem
that the frequency domain methods yield.

Keywords: Observer design, State variable feedback, Modern control, Matlab.

Introduction
The development of control system analysis and design can be divided into three eras. In the first era,
we have classical control theory, which deals with techniques developed before 1950. Classical control
embodies such methods as root locus, bode, Nyquist, and Routh-Hurwitz. These methods have in
common the use of transfer function in the complex frequency (s) domain, emphasis on the use of
graphical techniques, the use of feedback, and the use of simplifying assumptions to approximate the
time response.
In the second era, we have modern control, which refers to state space-based methods developed in the
late 1950s. and early 1960s. In modern control, system models are directly written in the time domain.
Analysis design are also done in the time domain. The period of 1960s was the heyday of modern
control. That period did not last very long, however. For one thing, classical control was already well
entrenched, tested, and established. Modern control provides a lot of insight into system structure and
properties that could be steadied manipulated using classical control. During the era of 1970s, and
1980s, a body of methods finally emerged that tried to provide answers to the plant uncertainty
problem.

FULL TEXT BOOK 530 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

These techniques, commonly Known as robust control, are a combination of modern state space and
classical frequency domain techniques.
The research highlights on state space design using MATLAB. State space design consist of
specifying the system's desired pole locations and then designing a controller consisting of state-
variable feedback gains to meet these requirements. If the state variables are not available, an observer
is designed to emulates the plant and provide estimated state variables.
The paper consist of four additional parts. In part 2, we give a brief presentation for analysis of control
system using state space. Part 3, is devoted for control system designing using state space. Controller
and observer design is reported in this part. In part 4, we use antenna azimuth position control system
as case study. The combined designed of controller and observer is demonstrated in this case study.
Part 5, gives the main conclusions and recommendations.
Stat-space analysis of control systems
Limitation of conventional control theory:
The root-locus method and frequency- response methods are quite useful for dealing whit single-input
single-output systems. For example, by means of open-loop frequency- response tests, we can predict
the dynamic behavior of the closed-loop system. If necessary, the dynamic behavior of a complex
system may be improved by inserting a simple lead or lag compensator. The technique of conventional
control theory are conceptually simple and require only a reasonable amount of computation.
In conventional control theory, only the input, and error signals are considered important; the analysis
and design of control systems are carried out using transfer functions, together with variety of
graphical techniques such as root-locus plots an Nyquist plots. The unique characteristic of
conventional control theory is that it is based on the input-output relation of the system, or the transfer
function.
The main disadvantage of conventional control theory is that, generally speaking, it is applicable only
to linear time-invariant systems having a single input and a single output. It is powerless for time-
varying systems, nonlinear system (except simple ones), and multiple input-multiple-output systems.
Thus conventional techniques (the root-locus and frequency-response methods) do not apply to the
design of optimal and adaptive control systems, which are mostly time varying and / or nonlinear.
Modern control theory
The modern trend in engineering system is towards greater complexity, due mainly to the requirements
of complex tasks and good accuracy. Complex system may have multiple inputs and multiple outputs
and may be time-varying. Because of the necessity of meeting increasingly stringent requirements on
the performance of control systems, the increase in system complexity, and easy access to large-scale
computers, modern control theory , which is a new approach to analysis and design of complex
control systems, has been developed since around 1960. This new approach is based on the concept of
stat. the concept of state by itself is not new since it has been in existence for a long time in the field of
classical dynamics and other fields. (in fact, the phase-plane method, is a two- dimensional state-space
method.)
Modern control theory versus conventional control theory
Modern control theory is contrasted whit conventional control theory in the former is graphical to
multiple- input multiple- output systems, which may be linear or nonlinear, time-varying while/ the
latter is applicable only to linear time-invariant single-input-single-output systems. Also, modern
control theory is essentially a time-domain approach, while conventional control theory is a complex
frequency- domain approach.
System design in classical control theory is based on trial-and-error procedures which, in general, will
not yield optimal control systems. System design in modern control theory, on the other hand, enables
the engineer to design optimal control systems whit respect to given performance indexes. In
additional, design in modern control theory can be carried out for a class of inputs, instead of a

FULL TEXT BOOK 531 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

specific input function, such as the impulse function, step function, or sinusoidal function. Also,
modern control theory enables the engineer to include initial conditions in the design.
Before we proceed further, we must define state, state variables, state vector, and state space.
State: the state of dynamic system is the smallest set of variables (called state variables) such that the
knowledge of these variables at t=t0, together whit the input for t≥t0 for completely determines the
behavior of the system for any time t≥t0.
Thus, the state of a dynamic system at time is uniquely determined by the state at time t 0 and the input
for t≥t0, and it is independent of the state and input before t0. Note that, in dealing with linear time-
invariant system, we usually choose the reference time t0 to be zero.
State variables: The state variables of a dynamic system are the smallest set of variables which
determine the state of the dynamic system. If at least n variables x1(t), x2(t),……,x(t) are needed to
completely describe the behavior of dynamic system (such that once the input is given for t≥t0 and the
initial state at t=t0 is specified, the future state of the system is completely determined), then such n
variables x1(t), x2(t),……,x(t) are a set of state variables. Note that the state variables need not be
physically measurable or observable quantities. Practically, however, ti is convenient to choose easily
measurable quantities for the state variables because optimal control laws will require the feedback of
all state variables with suitable weighting.
State vector: If n state variables are needed to completely describe the behavior of a given system,
then these n state variables can be considered to be the n components of a vector x(t). such a vector is
called a state vector. A state vector is thus a vector which determines uniquely the system state x(t) for
any t≥t0, once the input u(t) for t≥t0, is specified.
State space: The n-dimensional space whose coordinates axes consist of the x1 axis, x2 axis,….xn axis
called a state space. Any state can be represented by a point in the state space.
Analysis of complex systems
A modern complex system may have many inputs and many outputs, and these may be interrelated in
a complicated manner. To analysis such a system, it is essential to reduce the complexity of the
mathematical expression, as well as resort to computers for most of the tedious computations
necessary in the analysis. The state-space approach to system analysis is best suited from this
viewpoint.
While conventional control theory is based on the input-output relationship, or transfer function,
modern control theory is based on the description of system equations in terms of n first-order
differential equations. The use vector-matrix notation greatly simplifies the mathematical
representation of system of equations. The increase in the number of state variables, the number of
inputs, or the number of outputs does not increase the complexity of the equations. In fact, the analysis
of complicated multiple-input-output system can be carried out by the procedures that are only slightly
more complicated than those require for the analysis of system of first-order scalar differential
equations.
From the computational viewpoint, the state-space methods are particularly suited for digital-
computer computations because of their time-domain approach. This relieves the engineer of the
burden of tedious computation otherwise necessary and enables him to devote his effort solely to the
analytical aspect of the problem. This is one of the advantages of the state-space methods.
Finally, it is important to note that it is not necessary that the state variables represent physical
quantities of the system. Variables, which do not represent physical quantities, and those which are
neither measurable nor observable, may be chosen as state variable. Such freedom in choosing state
variable is another advantage of the state-space methods.
State –space representation of system
A dynamic system consist of a finite number of lumped elements may be described by ordinary
deferential equations in which time is the independent variable. By use of vector-matrix notation, an
nth-order differential equation may be expressed by a first-order vector-matrix differential equation. If

FULL TEXT BOOK 532 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

n elements of the vector are a set of state variables, then the vector-matrix differential equation is
called a state equation. In this section we shall present methods for obtaining state-space
representation of continuous- time system.
Consider the following nth-order system:
𝑦 (𝑛) + 𝑎1 𝑦 (𝑛−1) + ⋯ . 𝑎𝑛−1 𝑦̇ + 𝑎𝑛 𝑦 = 𝑢 (1)
𝑛−1 (0),
Noting that the knowledge of 𝑦(0), 𝑦̇ (0), … , 𝑦 together whit the input u(t) for t≥ 0, determines
completely the future behavior of the system, we may take 𝑦(0), 𝑦̇ (0), … , 𝑦 𝑛−1 (0), as a set of n state
variables. (Mathematically, such a choice of state variables is quite convenient. Practically, however,
because higher-order derivative terms inaccurate, due to noise effects inherent in any practical
situations, such a choice of the state variables may not be desirable.)
Let us define
X1 =y
X2 = 𝑦̇
…….
Xn =𝑦 (𝑛−1)
Then equation 1 can be written as
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 𝑛−1 = 𝑥𝑛
𝑥̇ 𝑛 = −𝑎𝑛 𝑥1−⋯……..− 𝑎1 𝑥𝑛 + 𝑢
Or
𝑥̇ = 𝐴𝑥 + 𝐵𝑢 (2)
Where:
𝑥1
𝑥2
𝑥= ⋮
⋮
[𝑥𝑛 ]
0 1 0 … 0 0
0 0 1 … 0 0
𝐴= ⋮ ⋮ ⋮ ⋮ ⋮ , 𝐵= ⋮
0 0 0 … 1 0
[−𝑎𝑛 −𝑎𝑛−1 −𝑎𝑛−2 … −𝑎1 ] [1]
The output equation becomes
𝑥1
𝑥2
𝑦 = [1 0 … 0] [ ⋮ ]
𝑥𝑛
Or
𝑦 = 𝐶𝑥 (3)
Where 𝐶 = [1 0 … 0]
The first-order differential equation, Eq.(2), is the state equation, and the algebraic Eq.(3), is the output
equation.

FULL TEXT BOOK 533 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Non-unique of the state variable

In has been stated that a set of state variables is not unique for a given system. Suppose that x1, x2 ….xn
are a set of state variables. Then we may take as another set of state variables any set of functions
𝑥̂1=X1(x1 , x2, ….,xn)
𝑥̂2=X2(x1 , x2, ….,xn)
………….
𝑥̂n=Xn(x1 , x2, ….,xn)
Provide that, for every set of values 𝑥̂1, 𝑥̂2 , …., 𝑥̂n , there corresponds a unique set of values x1 , x2,
….,xn and vice versa. Thus, if x is a state vector, then 𝑥̂ where 𝑥̂=Px is also a state vector, provided
the matrix P is nonsingular. Different state vectors convey the same information about the system
behavior.
Eigenvalue of an n X n matrix A
The eigenvalues of an n X n matrix A are the roots of characteristic equation
|𝜆I − A| = 0
The eigenvalues are sometimes called the characteristic roots.
Consider, for example, the following matrix A:
0 1 0
A=[ 0 0 1]
−6 −11 −6
The characteristic equation is
λ 1 0
|𝜆I − A| = [ 0 𝜆 1 ]
−6 −11 𝜆 + 6
=𝜆3 + 6𝜆2 + 11𝜆 +6
=(λ+1)(λ+2)(λ+3)=0
The eigenvalues of A are the roots of the characteristic equation, or -1 , -2, and -3.
Invariance of eigenvalues
The provide invariance of the eigenvalues under a linear transformation, we must show that the
characteristic polynomials
|𝜆I − A| and |𝜆I − 𝑃−1 A𝑃| are identical.
Since the determinant of product of the determinants, we obtain
|𝜆I − 𝑃 −1 A𝑃| = |𝜆I𝑃−1 𝑃 − 𝑃 −1 A𝑃|
=|𝑃−1 (𝜆I − A)𝑃|
=|𝑃−1 | |𝜆I − A| |𝑃|
=|𝑃−1 | |𝑃| |𝜆I − A|
Noting that the product determinants |𝑃−1 | and |𝑃| is the determinant of the product |𝑃−1 𝑃| ,we
obtain
|𝜆I − 𝑃−1 A𝑃| = |𝑃 −1 𝑃| |𝜆I − A|
= |𝜆I − A|
Thus we have proved that the eigenvalues of A are invariant a linear transformation.

FULL TEXT BOOK 534 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Diagonalization of n X n matrix
Note that of n X n matrix A whit distinct eigenvalues is given by

0 1 0 … 0
0 0 1 … 0
𝐴= ⋮ ⋮ ⋮ ⋮ ⋮ (4)
0 0 0 … 1
[−𝑎𝑛 −𝑎𝑛−1 −𝑎𝑛−2 … −𝑎1 ]
The transformation X= Pz where
1 1 … 1
𝜆1 𝜆2 … 𝜆𝑛
𝑃 = 𝜆12 𝜆22 … 𝜆2𝑛
⋮ ⋮ ⋮
𝜆1𝑛−1 𝜆𝑛−1
2 … 𝜆𝑛−1
𝑛
[ ]
λ1, λ2…..,λn= n distinct eigenvalues of A will transform P-1AP into the diagonal matrix, or
𝜆1 0
𝜆2
𝑃−1 A𝑃𝐼 = ⋱
⋱
[ 0 𝜆𝑛 ]
If the matrix A defined by Eq.4 involves multiple eigenvalues, then diagonalization is impossible, if
the 3X3 matrix A where
0 1 0
A=[ 0 0 1 ]
−𝑎3 −𝑎2 −𝑎1
Has the eigenvalues λ1, λ2, λ3, then the transformation x=Sz where
1 0 1
𝜆
S=[ 1 1 𝜆3 ]
𝜆12 2𝜆1 𝜆23
Will yield
𝜆1 1 0
−1
S AS = [ 0 𝜆1 0]
0 0 𝜆3
Such a form is called the Jordan canonical form.
State-space representation of nth-order system with r forcing functions
Consider the multiple-input-multiple-output system shown in Figure.2-1. In this system, x1,x2,……,xn
represent the state variables; u1, u2,…., ur denote the input variables; and y1, y2,……,ym are the output
variables. From figure.1, we obtain the system equations as follows:
𝑥̇ 1=𝑎11 (𝑡)𝑥1 + 𝑎12 (𝑡)𝑥2 + ⋯ . . +𝑎1𝑛 (𝑡)𝑥𝑛 + 𝑏11 (𝑡)𝑢1 + 𝑏12 (𝑡)𝑢2 + ⋯ + 𝑏1𝑟 (𝑡)𝑢𝑟
𝑥̇ 2=𝑎21 (𝑡)𝑥1 + 𝑎22 (𝑡)𝑥2 + ⋯ . . +𝑎2𝑛 (𝑡)𝑥𝑛 + 𝑏21 (𝑡)𝑢1 + 𝑏22 (𝑡)𝑢2 + ⋯ + 𝑏2𝑟 (𝑡)𝑢𝑟
𝑥̇ n=𝑎𝑛1 (𝑡)𝑥1 + 𝑎𝑛2 (𝑡)𝑥2 + ⋯ . . +𝑎𝑛𝑛 (𝑡)𝑥𝑛 + 𝑏𝑛1 (𝑡)𝑢1 + 𝑏𝑛2 (𝑡)𝑢2 + ⋯ + 𝑏𝑛𝑟 (𝑡)𝑢𝑟

FULL TEXT BOOK 535 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Figure1: Multiple-input- multiple-output.

Where the a(t)'s and b(t)'s are constants or functions of t. in terms of vector-matrix notation, these n
equations can be written compactly as
ẋ = A(t)x + B(t)u (5)
Where
𝑥1
𝑥2
x = [ ⋮ ] = state vector
𝑥𝑛
𝑢1
𝑢2
u = [ ⋮ ] = input(or control) vector
𝑢𝑛
𝑎11 (𝑡) 𝑎12 (𝑡) ⋯ 𝑎1𝑛 (𝑡)
𝑎21 (𝑡) 𝑎22 (𝑡) ⋯ 𝑎2𝑛 (𝑡)
A(t) = ⋮ ⋮ ⋮
𝑎𝑛1 (𝑡) 𝑎𝑛2 (𝑡) ⋯ 𝑎𝑛𝑛 (𝑡)
[ ]
𝑏11 (𝑡) 𝑏12 (𝑡) ⋯ 𝑏1𝑛 (𝑡)
𝑏21 (𝑡) 𝑏22 (𝑡) ⋯ 𝑏2𝑛 (𝑡)
B(t) = ⋮ ⋮ ⋮
𝑏𝑛1 (𝑡) 𝑏𝑛2 (𝑡) ⋯ 𝑏𝑛𝑛 (𝑡)
[ ]
Equation (5) is the state equation for the system. [Note that a vector-matrix differential equation such
as eq.(5)(or the equivalent n first-order differential equations) describing the dynamics of a system is
a state differential equation satisfies the definition of state variables.]
For the output signals, we obtain
𝑦1 =𝑐11 (𝑡)𝑥1 + 𝑐12 (𝑡)𝑥2 + ⋯ . . +𝑐1𝑛 (𝑡)𝑥𝑛 + 𝑑11 (𝑡)𝑢1 + 𝑑12 (𝑡)𝑢2 + ⋯ + 𝑑1𝑟 (𝑡)𝑢𝑟
𝑦2 =𝑐21 (𝑡)𝑥1 + 𝑐22 (𝑡)𝑥2 + ⋯ . . +𝑐2𝑛 (𝑡)𝑥𝑛 + 𝑑21 (𝑡)𝑢1 + 𝑑22 (𝑡)𝑢2 + ⋯ + 𝑑2𝑟 (𝑡)𝑢𝑟
ym=𝑐𝑚1 (𝑡)𝑥1 + 𝑐𝑚2 (𝑡)𝑥2 + … . . +𝑐𝑚𝑛 (𝑡)𝑥𝑛 + 𝑑𝑚1 (𝑡)𝑢1 + 𝑑𝑚2 (𝑡)𝑢2 + ⋯ + 𝑑𝑚𝑟 (𝑡)𝑢𝑟
In terms of vector-matrix notation, these m equations can be written compactly as:
y = C(𝑡)x + D(𝑡)u (6)
Where
𝑦1
𝑦2
y = [ ⋮ ]= output vector
𝑦𝑚

FULL TEXT BOOK 536 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

𝑐11 (𝑡) 𝑐12 (𝑡) ⋯ 𝑐1𝑛 (𝑡)

𝑐21 (𝑡) 𝑐22 (𝑡) ⋯ 𝑐2𝑛 (𝑡)
C(t) = ⋮ ⋮ ⋮
𝑐𝑛1 (𝑡) 𝑐𝑛2 (𝑡) ⋯ 𝑐𝑛𝑛 (𝑡)
[ ]
𝑑11 (𝑡) 𝑑12 (𝑡) ⋯ 𝑑1𝑛 (𝑡)
𝑑21 (𝑡) 𝑑22 (𝑡) ⋯ 𝑑2𝑛 (𝑡)
D(t) = ⋮ ⋮ ⋮
𝑑𝑚1 (𝑡) 𝑑𝑚2 (𝑡) ⋯ 𝑑𝑚𝑟 (𝑡)
[ ]
Equation (5) is the output equation for the system. Matrices A(t),B(t),C(t),and D(t) completely
characterize the system dynamics.
If the differential equation of the system involves derivatives of the forcing function, such as
𝑦 (𝑛) + 𝑎1 𝑦 (𝑛−1) + ⋯ + 𝑎𝑛−1 𝑦̇ + 𝑎𝑛 𝑦 = 𝑏0 𝑢(𝑛) + 𝑏1 𝑢(𝑛−1) + ⋯ + 𝑏𝑛−1 𝑢̇ + 𝑏𝑛 𝑢 (7)
Then the set of n variables 𝑦, 𝑦̇ , 𝑦̈ , … . , 𝑦 (𝑛−1) do not qualify as of state variables,

Figure 2: Block diagram representation of the system described by equations (5 and 6).
And the straightforward method previously employed cannot be used. This is because n first-order
differential equations.
𝑥̇ 1 = 𝑥2
𝑥̇ 2 = 𝑥3
𝑥̇ 𝑛 = −𝑎𝑛 𝑥1 −𝑎𝑛−1 𝑥2 − ⋯ −𝑎1 𝑥𝑛 + 𝑏0 𝑢𝑛 + 𝑏1 𝑢𝑛−1 + ⋯ + 𝑏𝑛 𝑢
Where x1= y may not yield a unique solution.
The main problem in defining the state variables for this case lies in the derivative terms on the right-
hand side of the last of the preceding n
The state variables must be such that they will eliminate the derivatives of u in the state equation.
It is a well-known fact in modern control theory we define the following n variables as a set of n state
variables
𝑥1 = 𝑦 − 𝛽0 𝑢
𝑥2 = 𝑦̇ − 𝛽0 𝑢̇ − 𝛽1 𝑢 = 𝑥̇ 1 − 𝛽1 𝑢
𝑥3 = 𝑦̈ − 𝛽0 𝑢̈ − 𝛽1 𝑢̇ − 𝛽2 𝑢 = 𝑥̇ 2 − 𝛽2 𝑢 (8)
𝑥𝑛 = 𝑦 𝑛−1 − 𝛽0 𝑢𝑛−1 − 𝛽1 𝑢𝑛−2 − ⋯ − 𝛽𝑛−2 𝑢̇ − 𝛽𝑛−1 𝑢 = 𝑥̇ 𝑛−1 − 𝛽𝑛−1 𝑢
Where β, β0, β1, β2 ,……. Βn are determined from
β0=b0
β1=b1-a1 β0

FULL TEXT BOOK 537 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

β2=b2-a1 β1- a2 β0
β3=b3-a1β2-a2β1-a3β0 (9)
βn=bn-a1 βn-1- …..-an-1β1-anβ0
Then the existence and uniqueness of the solution of the state equation is guaranteed.(note that this is
not only choice of state variables, we obtain the following state equation and output equation for the
system of Eq.(6)
𝑥̇ 1 0 1 1 ⋯ 0 𝑥1 β1
𝑥̇ 2 0 0 1 ⋯ 0 𝑥2 β2
⋮ = ⋮ ⋮ ⋮ ⋯ ⋮ ⋮ + ⋮ [u]
𝑥̇ 𝑛−1 0 0 0 ⋯ 1 𝑥𝑛−1 β𝑛−1
[ 𝑥̇ 𝑛 ] [ −𝑎𝑛 −𝑎𝑛−1 −𝑎𝑛−1 ⋯ −𝑎1 ] [ 𝑥𝑛 ] [ β𝑛 ]
𝑥1
𝑥2
𝑦 = [1 0 ⋯ 0] ⋮ + β0 u
𝑥𝑛−1
[ 𝑥𝑛 ]
Or ẋ = Ax + Bu (10)
𝑌 = Cx + Du (11)
Where
𝑥1
𝑥2
X= ⋮ ,
𝑥𝑛−1
[ 𝑥𝑛 ]
0 1 1 ⋯ 0 β1
0 0 1 ⋯ 0 β2
A= ⋮ ⋮ ⋮ ⋯ ⋮ ,B= ⋮
0 0 0 ⋯ 1 β𝑛−1
[ −𝑎𝑛 −𝑎𝑛−1 −𝑎𝑛−1 ⋯ −𝑎1 ] [ β𝑛 ]
C = [1 0 ⋯ 0], D = β0 = 𝑏0
The initial condition x(0) may be determined by use of Eq.(8).
In this state-space representation, the matrix A is essentially the same as in the system of Eq (1). The
derivatives on the right-hand side of Eq.(6)
Affect only the element of the B matrix.
Note that state-space representation for the following transfer function
𝑌(𝑠) 𝑏0 𝑠 𝑛 + 𝑏1 𝑠 𝑛−1 + ⋯ + 𝑏𝑛−1 𝑠 + 𝑏𝑛
= 𝑛
𝑈(𝑠) 𝑠 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑠 + 𝑎𝑛
Is given also by Eqs. (10) and (11).
Transfer function matrix
The concept of the transfer matrix is an extension of the transfer function; here we shall first obtain
transfer functions of single-input-single-output system and then transfer matrices of multiple-input-
multiple-output system from state and output equations.
We shall derive transfer function of a single-input-single-output system from the Laplace-transformed
version of the state and output equations.
Let us consider the system whose transfer function is given by

FULL TEXT BOOK 538 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

𝑌(𝑠)
𝑈(𝑠)
= 𝐺(𝑠) (12)
we showed that the state-space representation for this system is given by
ẋ = Ax + Bu (13)
𝑌 = Cx + Du (14)
Where X is the state vector, u is the input, and y is the output. The Laplace transforms of Eq (13) is
given by
sX(s)-X(0)=AX(s)+Bu(s) (15)
Y(s)=CX(s)+Du(s) (16)
Since the transfer function was previously defined as ratio of Laplace transform of the output to the
Laplace transform of the input when the initial conditions were zero, we assume that X(0).
In Eq.(15) is zero. By substituting
X(s)=(sI-A)-1BU(s) into Eq.(16)
We obtain
Y(s)=[C(sI-A)-1B+D]U(s) (17)
Upon comparing Eq.(17)with Eq(12), we see that
G(s)=C(sI-A)-1B+D (18)
This is the transfer-function expression in terms of A,B,C,and D.
Note that the right-hand side of Eq.(18) involves (sI-A)-1. Hence G(s) can be written as
𝑄(𝑠)
G(s) =
|SI − A|
Where 𝑄(𝑠) is polynomial in s. there for, |SI − A| is equal to the characteristic polynomial of G(s). In
the words, the eigenvalues of A are identical to the poles of G(s).
The transfer matrix G(s) relates the output Y(s) to the input U(s), or
Y(s)=G(S)U(s) (19)
If the input vector U is r dimensional and the output y is m dimensional, then the transfer matrix is an
mXr matrix. In an expanded form, Eq(19) can be written as
𝑦1 𝐺11 (𝑠) 𝐺12 (𝑠) ⋯ 𝐺1𝑟 (𝑠) 𝑢1
𝑦2 𝐺 (𝑠) 𝐺22 (𝑠) ⋯ 𝐺2𝑟 (𝑠) 𝑢2
[ ⋮ ] = [ 21 ][ ⋮ ]
⋮ ⋮ ⋮ ⋮
𝑦𝑚 𝐺𝑚1 (𝑠) 𝐺𝑚2 (𝑠) ⋯ 𝐺𝑚𝑟 (𝑠) 𝑢𝑚
The (i,j)th element Gij(s) of the transfer function relating the ith output to the jth. By following the
same steps as used in deriving Eq (18)
We obtained the transfer matrix for multiple-input-multiple-output system as follow:
G(s)=C(sI-A)-1B+D
Clearly, the transfer function expression given by Eq.(18) is a special case of this transfer matrix
expression.
Transfer matrix of closed-loop systems
Consider the system shown in figure(3). The system has multiple inputs and multiple outputs. The
transfer matrix of the feed forward path is G0(s), and that of the feedback path H(s). The transfer
matrix between the feedback signal vector E(s) is obtained as follows: since
B(s)=H(s)Y(s)

FULL TEXT BOOK 539 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

=H(s)G0(s)E(s)

Figure 3: block diagram

We obtain the matrix transfer matrix between B(s) and E(s) to be H(s)G0(S) thus the transfer matrix of
the cascade elements is the product of the transfer matrices of the individual elements. (Note that the
order of the matrix multiplication is very important since matrix multiplication is, in general, not
commutative.)
The transfer matrix of the closed-loop system is obtained as follows: since
Y(s) = G0(s)[U(s)-B(s)]
= G0(s)[U(s)-H(s)T(s)]
We obtain
[I+G0(s)H(s)]Y(s) = G0(s)U(s)
Pre multiplying both sides of this last equation by [I + G0(s)H(s)]-1, we obtain
Y(s) = [I + G0(s)H(s)]-1 G0(s)U(s)
The closed-loop transfer matrix G(s) is given by
G(s)=[I+G0(s)H(s)]-1G0(s) (20)
State-space representation of discrete-time system.
The state-space approach to the analysis of dynamic system can be extended to the discrete-time case.
The discrete form of the state-space representation is quite analogous to the continuous form to the
continuous form . the most general state-space representation of linear discrete-time systems is
X(k+1G(k)x(kH(k)u(k) (21)
Y(k)=C(k)x(k)+D(k)u(k) (22)
Where x(k) is the state vector, u(k) is the input vector, and y(k) is the output vector, each specified at t
= kT, k = 0,1,….., and T is sampling period.
Similarly, we use the simplified notation u(k), y(k),G(k),H(k), and D(k). Equation (21)and (22)
discrete-time system described by Eqs. (21)and (22).the unit-delay element has a delay time of T
seconds. If the linear discrete-time system is time-invariant, then Eqs.(21) and (22) are modified to
X(k+1)=Gx(k)+Hu(k) (23)
Y(k)=Cx(k)+Du(k) (24)
In this section, we shall be concerned mostly with systems described by Eqs.(23)and(24).

FULL TEXT BOOK 540 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Figure 4: block diagram representation of the discrete-time system described by Eqs.(21)and(22).

Discretization of continuous-time state space equations
If we wish to compute the state x(t) by use of a digital computer, we must convert a continuous-time
state equation to a discrete-time state . in the following, we shall present such a procedure. We assume
that the input vector u(t) changes only at the quality spaced sampling instants.( the sampling operation
here is fictitious. We shall derive the discrete-time state equation which yields the exact values at t =
kT, k = 0,1,2,….)
Consider the continuous-state equation
ẋ = Ax + Bu (25)
In the following, in order to clarify the analysis, we shall use the notation kT and (k+1) instead of k and
k+1. The discrete-time representation of Eq. (12) will take the form
x((k+1)T=G(T)x(kT)+H(T)u(kT) (26)
Note that the matrices G and H depend on the sampling period T.
(Once this period is fixed, g and H are constant matrices.)
In order to determine G(T) and H(T), we use the solution of Eq. (25),or
𝑡
𝐴𝑡 𝐴𝑡
x(t) = 𝑒 x(0) + 𝑒 ∫ 𝑒 −𝐴𝜏 Bu(τ) dτ
0
we assume that all the components of u(t) are constant over the interval between any two consecutive
sampling instants, or u(t) = u(kT) for the kth sampling period. Since
(𝑘+1)𝑇
x((𝑘 + 1)𝑇) = 𝑒 𝐴(𝑘+1)𝑇 x(0) + 𝑒 𝐴(𝑘+1)𝑇 ∫ 𝑒 −𝐴𝜏 Bu(τ) dτ (27)
0

And
𝑘𝑇
x(kt) = 𝑒 𝐴𝑘𝑇 𝑥(0) + 𝑒 𝐴𝑘𝑇 ∫0 𝑒 −𝐴𝜏 𝐵𝑢(τ) dτ (28)
Multiply Eq.(28) by eAT and subtraction from Eq.(27) we obtain
(𝑘+1)𝑇
x((k + 1)𝑇) = eAT x(𝑘𝑇) + e 𝐴(𝑘+1)𝑇 ∫ 𝑒 −A𝜏 B𝑢(τ) 𝑑𝜏
kT
AT AT 𝑇
=e x(𝑘𝑇) + e ∫0 𝑒 −A𝜏 B𝑢(𝑘𝑇) 𝑑𝑡
𝑇
=eAT x(𝑘𝑇) + ∫0 𝑒 A𝜆 B𝑢(𝑘𝑇) 𝑑𝜆 (29)
Where 𝜆 =T-t. if we define
G(T)= eAT (30)
𝑇
H(T)=(∫0 𝑒 A𝜏 𝑑𝑡)B (31)
Then Eq.(29) becomes
X((k+1)T) = G(T)x(kT)+H(T)u(kT)
Which is Eq.(26). Thus Eqs. (30)and (31) give desired matrices G(T) and H(T).
Design via state space.
Controller design
In this section we show how to introduce additional parameters into a system so that we can control
the location of all closed-loop poles. As nth-order feedback, control system has an nth-order closed-
loop characteristics equation of the form

FULL TEXT BOOK 541 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Sn+an-1Sn-1+…+a1S+a0=0 (32)
Since the coefficient of the highest power of S is unity, there are n coefficients whose values
determine the system's closed-loop pole locations. Thus, if we can introduce n adjustable parameters
into the system and relate them to the coefficients in Eq(32) , all of the poles of the closed-loop system
can be set to any desired location.
Topology for pole placement
In order to lay the groundwork for the approach, consider a plant represented in state-space by
ẋ = Ax + Bu (33a)
𝑌 = Cx (33b)
And shown pictorially in figure (5-a), where light lines are scalars and the heavy lines are vectors.
In a typical feedback control system the output, y , is feedback to the summing junction. It is
now that the topology of the design changes. Instead of feeding back.

(5-a)

(5-b)
Figure 5: a. state-space representation of plant b: plant with state-feedback.
y, what if we feedback all of the state variables if each state variable is feedback to control, u, through
a gain m ki, there would be n gains, ki, that could be adjusted to yield the required closed-loop pole
values. The feedback through the gains, ki, is represented in Figure(5-b) by the feedback vector –k.
the state equations for the closed-loop of figure 5-a can be written by inspection as:
ẋ = Ax + B𝑢 =Ax+B(-Kx+r)
=(ABK)x+Br (34.a)
y=Cx (34.b)
Before continuing you should have a good idea of how the example, assume a plant signal-flow graph
in phase-variable form, as shown in Figure (6.-a). Each state variable is then feedback to the plant's
input, u, through a gain, ki, as shown in Figure (6-b). The phase-variable form, with its typical lower
companion system matrix, or the controller canonical form, with its typical upper companion system
matrix, yields the simplest evaluation of the feedback gains. In our discussion, we use the phase-
variable form to develop and demonstrate the concepts.
The design of state-variable feedback for closed-loop pole placement consists of equating the
characteristic equation of a closed-loop system, such as that shown in Figure (6-b), to a desired
characteristic equation and then finding the values of the feedback gains, ki,

FULL TEXT BOOK 542 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

If a plant like that shown in Figure (6-b) is of high order and not represented in phase-variable or
controller canonical form, the solution for the ki's, can be intricate. Thus, it is advisable to transform
the system to either of these forms, design the ki's, and then transform the system back to its original
representation. We perform

(a)

(b)
Figure (6): a. phase-variable representation for plant; b. plant with state-variable feedback.

Pole placement for plants in phase-variable form

To apply pole-placement methodology to plants represented in phase-variable form, we take the
following steps:
1. represent the plant in phase variable form.
2. feedback each phase variable to the input of the plant through a gain, ki.
3. find the characteristic equation for the closed-loop system represented in step2.
4. decide upon all closed-loop locations and determine as equivalent characteristic equations.
5. equate like coefficients of the characteristic equations from step3 and 4 and solve for ki.

Controllability
To control the pole location of the closed-loop system, we are saying implicitly that the control signal,
u, can control the behavior of each state variable in x. if any one of the state variables cannot be
controlled by the control u, then we cannot place the poles of the system where we desire.
Observer Design
Control design relies upon access to the state variables for feedback through adjustable gains. This
access can be provided by hardware. For example, gyros can measure position and velocity on a space
vehicle. Sometimes it is impractical to use this hardware for reasons of cost, accuracy, or availability.
For example, in powered flight of space vehicles, inertial measuring units can be used to

FULL TEXT BOOK 543 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

(a)

(b)

(C)
Figure 7: state-feedback design using as observer to estimate unavailable state variable, a: open-loop
observer, b: closed-loop observer: exploded view of a close-loop observer, showing feedback
arrangement to reduce state variable estimation error.

Calculate the acceleration. However, their alignment deteriorates with time; thus, other means of
measuring acceleration may be desirable. In other applications, some of the state variables may not be
available at all, or it is too costly to measure them or send them to the controller. If the state variables
are not available because of system configuration or cost, it is possible to estimate the states.
Estimated states, rather than actual state, are then fed to the controller. One scheme is shown Figure
(6-a). an observer, sometimes called an estimator, is used to calculate state variables that are not
accessible form the plant. Here, the observer is a model of the plant. Let us look at disadvantages of
such a configuration.
Observability
Recall that the ability to control all of the state variable s is requirement for the design of controller.
State-variable feedback gains cannot be designed if any state variable is uncontrollable. Uncontrollable
can be viewed best with diagonal zed systems. The signal-flow graph showed clearly that the
uncontrollable sate variable was not connected to the control signal of the system.
Antenna control: design of controller and observer

FULL TEXT BOOK 544 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

In this case study we will use antenna azimuth position control system to demonstrate the combined
design of a controller and observer.
The block diagram of the original system is shown in Figure (7) arbitrarily setting the amplifier gain to
200. The block diagram parameters are kpot= 0.318, k1=100, a=100,km=2.083, am=1.71,kg=0.1.
The open-loop transfer function is
Y(s) 1325
G(𝑠) =
U(s) 𝑠(𝑠 + 1.71)(𝑠 + 100)
The output of the system is azimuth position θ0(s).

Figure 8: Block diagram of antenna a zimuth position control system.

The case study will specify a transient response for the system and a faster transient response for the
observer. The final design configuration will consists of the plant, the observer, and the controller as
show in Figure (9). The design of the observer and controller will be separate.

Figure 9: state-space design configuration shown plant, observer, and controller plant.

Design specification
We will design a controller to yield a10percentage overshoot and a settling time of 1second. Place the
third pole 10 times father from the imaginary axis than the second order dominant pair.
We will design an observer to estimate the states. The desired transient response for the observer is
10% overshoot and a natural frequency 10times greater than the system response above.
An in the case of the controller, we will place the third pole 10 times farther from the imaginary axis
than observer's dominant second-order pair.
Controller design

FULL TEXT BOOK 545 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

We first design the controller by finding the desired characteristic equation. A 10% overshoot and
settling time of 1 second yields =0.591 and n=6.67rad/sec. thus the dominant poles are located as -
4j5.46 . The desired characteristics equation for the closed loop system is
(s2+8s+45.8)(s+40)=s3+48s2+365.8s (35)

Next, we will find the actual characteristic equation of the closed loop system. The first step is to
model the closed-loop system in state space and then find it is characteristic equation.
The plant is:
1325 1325
𝐺(𝑠) = 𝑠(𝑠+1.71)(𝑠+100) = 𝑠(𝑠2 +101.7𝑠+171) (36)

Using state variable, this transfer function is converted to state-space from using MATLAB statement
tfss:

−101.71 −171 0 0
ẋ = [ 1 0 0] x + [0] u (37)
0 1 0 1
𝑦 = [0 0 1325]x (38)
We now pause in our design to evaluate controllability of the system. The controllability matrix, Cm, is

0 0 1
𝐂𝐌 = [0 1 −201.71 ] (39)
1 −101.71 10173.92
Noting that the matrix CM is obtained using MATLAB statement Cm=ctrb(A,B). the determinate of
CM is -1; thus, the system is controllable.
Continuing with the design of the controller, we now find the characteristic equation of the system.
The system matrix, A-BK,
0 1 0
𝐀 − 𝐁𝐊 = [ 0 0 1 ] (40)
−𝑘1 −(171 + 𝑘2 ) −(101.71 + 𝑘3 )

Thus, the closed-loop system’s characteristic equation is

det[sI-(A-BK)]= s3+(101.71+k3)s2+(171+k2)s+k1=0 (41)

We evaluate the controller gain matrix K. we use the MATLAB command K=acker(A,B,poles), which
calculate the controller gain from the system matrix (A) and the input matrix(B):
K=[1832 194.8 -53.71] (42)

Observer Design.
Before designing the observer, we test the system observability. Using the A and C matrices, the
observability matrix, OM, is
1325 0 0
𝐎𝐌 = [ 0 1325 0 ] (43)
0 0 1325

FULL TEXT BOOK 546 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

The observability matrix is obtained using MATLAB command Om=obsv(A,C).

det(OM)=13253
Thus OM is of rank 3, and the system is observable.
We now procced to design the observer, we need first to find (A-LC), where
𝑙1
L = [𝑙2 ] (44)
𝑙3
−1325𝑙1 1 0
A − LC = [−1325𝑙2 0 1 ] (45)
−1325𝑙3 −171 −101.71
The characteristic equation for the observer is now evaluated as
det=[λI-(A-LC)]
=λ3+(1325l1+101.71)λ2(134800l1+1325l2+171) λ+(22600l1+134800l2+1325l3)=0 (46)
From the problem statement, the poles of the observer are to be placed to yield a 10% overshoot and a
natural frequency 10 times that of the system's dominant poles yield. Thus the observer's dominant
poles yield (s2+80s+4583).
The real pant of the roots of this polynomials is -40. The third pole is then placed 10 times farther from
the imaginary axis at -400. The characteristic equation for the observer is
(s2+80s+4583)(s+400)=s3+480s2+36583s+1833200=0 (47)
We can obtain the observer gain matrix L. we use the MATLAB command l=acker(A',C',poles)' to
calculate the observer gain matrix L:
0.286
L=[−1.57] (48)
1494
Figure (9), which follows the general configuration of figure (9). Shows the completed design,
including the controller and the observer. The results of the design are shows in Figure (10) and
Figure (11). Figure (10) shows of the impulse response of the closed loop system. The undershoot and
settling time approximately meet the requirements set forth in the problem statement of 10% and 1
second, respectively. Figure (11) shows the error between the actual plant output y and the estimated
output 𝑦̂. It is shown that this error reach zero very fast.

Figure 10: completed state-space design for the antenna azimuth position control system showing
controller and observer.

FULL TEXT BOOK 547 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

Figure 11: designed impulse response of antenna azimuth position control system.

Figure 12: actual plant output and between (𝑦 − 𝑦̂).

Conclusion
Controller design via state-space techniques consist of feeding back the state variables to input, u, of
the system through specified gains. In some cases the control signal cannot effect on wore state
variables. For this system, a total design is not possible. Using the controllability matrix, a designer
can tell whether a system is controllable prior design.
Observer design consist of feeding back the error between the actual output and the estimated output.
This error is feedback through specified gains to the derivatives of the estimated state variables. The
response of the observer is designed to be faster than that of the instantaneously at the controller. For
systems, the observer requires the state variables cannot be deduced from the output of the system, as.
Using the observability matrix, the designer can ten whether or not a system is observable. Observes
can be designed only for observable systems.
Three advantages of state-space design are apparent. First, in contrast to the root locus method, all pol
locations can be specified to ensure a negligible effect of the non-dominant poles upon the transient
response. Second, with the use of an observer, we are no longer forced to acquire the actual system
variables for feedback. The advantage here is that sometimes the variables cannot by physically
accessed, or it may be too expensive to provide that access. Finally, the methods shown lend
themselves to design automation using the digital computer.

FULL TEXT BOOK 548 www.izdas.org

 AL-FARABI INTERNATIONAL CONGRESS ON APPLIED SCIENCES - II
May 2-4, 2021 / "Nakhchivan" University, Azerbaijan

References
Brogan, W.L, Modern control theory, Englewood Ccliffs, N.J : 3rd edition, prentice Hall, 199.
Decalo, R. Linear Systems: A state variable Approach with Numerical Implementation, Prentice Hall,
1989.
Tstefani, R.T, Design of Feedback control systems, Saunder colleg Publishing, 1994.
Bode,H.W., Network Analysis and Feedback Design, New York:Van Nostrand Reinhold,1945.
Anderson, B. D. O., and J. B. Moore, Linear Optimal Control, Upper Saddle River, NJ: Prentice Hall,
1971.
Cheng, D. K., Analysis of Linear Systems. Reading, MA: Addison- Wesley Publishing Company, Inc.,
1959.
A. A. Ahmed and A. F. Saleh Alshandoli, "On replacing a PID controller with Neural Network
controller for Segway," 2020 International Conference on Electrical Engineering (ICEE), 2020, pp. 1-
4, doi: 10.1109/ICEE49691.2020.9249811.
E. Kim, C. Kim and H. Shin, "Nonlinear State Feedback Controller Design for Buck-Boost
Converters," 2018 21st International Conference on Electrical Machines and Systems (ICEMS), 2018,
pp. 2418-2422, doi: 10.23919/ICEMS.2018.8549320.
X. Liu, Q. Zhu, P. Narayan and Y. Yang, "The first stage studies of U-state space control system
design," 2015 34th Chinese Control Conference (CCC), 2015, pp. 946-951, doi:
10.1109/ChiCC.2015.7259761.
M. M'Saad and M. Farza, "Output feedback control of a class of nonlinear systems," CCCA12, 2012,
pp. 1-2, doi: 10.1109/CCCA.2012.6417892.
J. Kim, G. Park and H. Choi, "Sliding mode control design under partial state feedback for ball and
beam system," ICCAS 2010, 2010, pp. 1293-1296, doi: 10.1109/ICCAS.2010.5670233.
M. C. Algrain, "Design of observers and feedback controllers using pseudo-linear models of nonlinear
systems," Proceedings of 1995 American Control Conference - ACC'95, 1995, pp. 3714-3718 vol.5,
doi: 10.1109/ACC.1995.53383.
G. Mingde and C. Qiusheng, "Position controller based on state observer for 6-DOF parallel
teleoperation manipulator," Proceedings of 2011 International Conference on Electronic & Mechanical
Engineering and Information Technology, 2011, pp. 1787-1790, doi: 10.1109/EMEIT.2011.6023450.
S. Moon, J. Choe and J. Lai, "Design of a state-space controller employing a deadbeat state observer
for ups inverters," 2017 Asian Conference on Energy, Power and Transportation Electrification
(ACEPT), 2017, pp. 1-6, doi: 10.1109/ACEPT.2017.8168593.
A. A. Ahmed, J. Santhosh and F. W. Aldbea, "Vehicle Dynamics Modeling and Simulation with
Control Using Single Track Model," 2020 IEEE International Women in Engineering (WIE)
Conference on Electrical and Computer Engineering (WIECON-ECE), 2020, pp. 1-4, doi:
10.1109/WIECON-ECE52138.2020.9397983.
I. McInerney and E. C. Kerrigan, "Automated Project-Based Assessment in a Predictive Control
Course," 2018 UKACC 12th International Conference on Control (CONTROL), 2018, pp. 443-443,
doi: 10.1109/CONTROL.2018.8516728.
T. Vámos, L. Keviczky, R. Bars, A. Benedek and D. Sik, "An Introductory Overview About Systems
and Control a Motivation Lecture in Control Education," 2018 26th Mediterranean Conference on
Control and Automation (MED), 2018, pp. 1-6, doi: 10.1109/MED.2018.8442604.
M. C. M. Teixeira, "Direct expressions for Ogata's lead-lag design method using root locus," in IEEE
Transactions on Education, vol. 37, no. 1, pp. 63-64, Feb. 1994, doi: 10.1109/13.275189.

FULL TEXT BOOK 549 www.izdas.org

View publication stats