Review Paper

REVIEW PAPER
"Enlarging the Domain of Attraction in Nonlinear Systems with Input Saturation"

E. Andry Dwi Kurniawan (20123007)
NONLINEAR CONTROL THEORY
E. Andry Dwi Kurniawan (20123007)
The introduction of the paper "Enlarging the domain of attraction in nonlinear systems
with input saturation" provides a comprehensive overview of the significance of input
saturation in dynamical control systems and its impact on system stability. Here is a summary
of the key points covered in the introduction:
1. Importance of Input Saturation: Input saturation is highlighted as a common
phenomenon in dynamical control systems that can degrade system performance and
lead to instability. Researchers have focused on studying the effects of saturation on
system behavior and stability.
2. Stability Analysis: The introduction emphasizes the critical role of stability in system
engineering, particularly in the analysis of nonlinear dynamic systems. Stability is
crucial for ensuring that a system returns to its equilibrium position when perturbed,
and the domain of attraction plays a key role in assessing system stability.
3. Domain of Attraction: The concept of the domain of attraction is introduced as the
maximum distance between the initial state and the equilibrium position of a system.
Estimating and evolving the domain of attraction have been challenging tasks that have
attracted the interest of researchers in recent decades.
4. Methods for Estimating Domain of Attraction: The introduction mentions the use of
Lyapunov and non-Lyapunov methods for estimating the domain of attraction in
nonlinear systems. Lyapunov functions are highlighted as essential tools for analyzing
and controlling system stability.
5. Research Focus: The introduction sets the stage for the study by highlighting the
unresolved challenges in estimating and evolving the domain of attraction in nonlinear
systems with input saturation. The authors aim to address these challenges by proposing
a methodology that combines stability analysis with the JAYA optimization algorithm
to maximize the domain of attraction
Overall, the introduction provides a solid foundation for understanding the motivation behind
the study, the importance of stability in nonlinear systems, and the significance of the domain
of attraction in assessing system performance and stability.
The section on "Problem Formulation and Preliminaries" in this paper lays the
groundwork for the study by defining the nonlinear system with input saturation and
introducing key concepts related to stability analysis and the domain of attraction. Here is a
summary of the key points covered in this section:
1. Nonlinear System with Input Saturation: The section introduces the nonlinear
system with input saturation, represented by a set of differential equations that include
the effects of saturation on the system dynamics. The system is formulated to include
an equilibrium point at the origin and a control input parameter K that influences system
stability and the region of attraction.
Consider the following nonlinear system with input saturation:
𝑑𝑥
{ 𝑑𝑡 = 𝑓(𝑥) + 𝐺𝑠𝑎𝑡(𝑢) (1)
𝑢 = 𝑘𝑥
function: 𝑠𝑎𝑡(𝑢) = 𝑠𝑖𝑔𝑛(𝑢)𝑚𝑖𝑛{|𝑢|, 1}.
𝑢 = 𝐾𝑥 (2)
2. Differential Algebraic Representation (DAR): The DAR is utilized to represent the

nonlinear system with input saturation in a structured form that accounts for the
nonlinear dynamics and the presence of saturation effects. This representation
facilitates the analysis of system behavior and the application of optimization
techniques to maximize the region of attraction.
There many representation of the system nonlinear which a dynamical system (1) can
be represented in the following differential Algebraic representation (DAR);
𝑑𝑥
= 𝐴1 𝑥+ 𝐴2 𝜋 + 𝐴3 𝑠𝑎𝑡(𝑢)
{ 𝑑𝑡 (3)
0 = Υ1 𝑥 + Υ2 𝜋 + Υ3 𝑠𝑎𝑡(𝑢)
Where 𝜋 is auxiliary nonlinear vector function 𝑥, 𝑠𝑎𝑡(𝑢) containing rational and
polynomial terms. 𝐴1 , 𝐴2 and 𝐴3 are constant matrices. Υ1 ∈ ℜ 𝑧×𝑛 , Υ2 ∈ ℜ𝑛𝑧 ×𝑛𝑧 and
Υ3 ∈ ℜ1×𝑛𝑧 are affine matrix functions of 𝑥. The matrix Υ2 ∈ ℜ𝑛𝑧 ×𝑛𝑧 has full coloumn-
rank. If Υ2 is invertible, as according to (3) we have that
𝜋(𝑥, 𝑠𝑎𝑡(𝑢)) = −Υ2−1 (Υ1 𝑥(𝑡) + Υ3 𝑠𝑎𝑡(𝑢))
Let consider the following dead zone non-linearity
𝜓(𝑢) = 𝑠𝑎𝑡(𝑢) − 𝑢 (4)
3. Optimization Objective: The primary objective of the study is to find the optimal
value of the control input parameter K that ensures the system's asymptotic stability at
the origin while maximizing the region of attraction. This optimization task is essential
for enhancing system performance and stability in the presence of input saturation.
4. Importance of LMI Conditions: The section highlights the significance of Linear
Matrix Inequality (LMI) conditions in developing the optimization approach and
leveraging the Lyapunov theory for stability analysis. The LMI conditions play a crucial
role in determining the optimal control input parameter and expanding the domain of
attraction.
The "Stability Analysis" section focuses on the development of stability conditions using
the theory of Lyapunov. Here is a summary of the key points covered in this section:
1. Lyapunov Theory: The section introduces the theory of Lyapunov as the basis for
stability analysis in nonlinear systems. It emphasizes the importance of Lyapunov
functions in characterizing the stability of systems and estimating the domain of
attraction.
2. Domain of Attraction Definition: The domain of attraction is defined as the set of
initial states from which the system trajectories converge to the equilibrium point. The
section presents the domain of attraction as a positive set where trajectories approach
the origin as time approaches infinity.
Following the theory of Lyapunov, we can present the domain of attraction in which
the stability of systems defined as it represents by the following set:
𝜙𝑉 = {𝑥 ∈ ℜ𝑛 : 𝑉(𝑥) ≤ 1}
3. Quadratic Lyapunov Function: A candidate Lyapunov function of quadratic type is
introduced for stability analysis. The quadratic Lyapunov function is defined using a
positive symmetric constant matrix to assess system stability and convergence to the
equilibrium point. In this work, we introduced the candidate Lyapunov functions of
quadratic type:
𝑉(𝑥) = 𝑥 𝑇 𝑃𝑥
With, 𝑃 = 𝑃𝑇 is a positive symmetric constant matrix.
4. Derivative Analysis: The section presents the derivative of the quadratic Lyapunov
function in matrix form to analyze the system's stability properties. The derivative is
used to establish conditions for system stability and convergence to the equilibrium
point.
5. Stability Conditions: Stability conditions based on the Lyapunov theory are derived
to ensure system stability and convergence. These conditions involve analyzing the
derivative of the Lyapunov function and verifying specific relationships to guarantee
stability.
The "Enlarging the Domain of Attraction" section focuses on techniques to expand the
region of attraction for nonlinear systems subject to input saturation. Here is a summary of the
key points covered in this section:
1. Objective: The primary objective of this section is to enhance the existing technique
by optimizing the gain using a JAYA optimization algorithm to maximize the region of
attraction (ROA) for nonlinear control systems with input saturation represented in
algebraic differential form. Additionally, a generalized sectorial condition is proposed
to mitigate the effects of saturation.
The following set 𝜙𝑉 of the Lyapunov function 𝑉(𝑥) well introduced by;
𝜙𝑉 = {𝑥 ∈ ℜ𝑛 : 𝑉(𝑥) ≤ 1}
Then 𝜙𝑉 is said the DA of the origin for (1). In order to find the maximal DA, it needed
to find the maximum value of K (the input control) for 𝜙𝑉 (𝐾) such that the calculi set
satisfies the hypothesis of the Lyapunov theory.
So, the largest region of Attraction approximate for the origin of system (1) described
by:
𝐾 ∗ = max 𝐾
2. Techniques: The section discusses the utilization of Lyapunov theory-based
techniques and the concept of Lyapunov functions to estimate the region of attraction.
Various types of Lyapunov functions are mentioned, including polynomial, quadratic,
ellipsoidal, rational, and piecewise quadratic functions.
3. Approach: The approach involves analyzing stability using Lyapunov theory,
integrating the JAYA optimization algorithm to maximize the domain of attraction, and
implementing a generalized sectorial condition to address saturation effects.
The performance of the algorithm proposed is quite
satisfactory for the reason that it converges to the optimal in
less than 10 iteration as shown in the figure (1).
Figure4: Result of input control of example without JAYA.
Figure 1:Dynamic of K using the JAYA approach.
Figure5: Result of states of example with input control

optimize.
Figure 2: Estimates of the region of attraction of system

(33)such that the red domain of attraction correspondent to
without JAYA approach and the blue ellipsoid with JAYA.
Figure6: Control input optimize.

To show the performance of the optimization algorithm
JAYA, we want to repeat the same simulation of the
example.
Figure3: Result of states of example without JAYA.

Review Paper - Nonlinear Control Theory - E. Andry Dwi Kurniawan (20123007)

Uploaded by

Review Paper - Nonlinear Control Theory - E. Andry Dwi Kurniawan (20123007)

Uploaded by

"Enlarging the Domain of Attraction in Nonlinear Systems with Input Saturation"

2. Differential Algebraic Representation (DAR): The DAR is utilized to represent the

Figure4: Result of input control of example without JAYA.

Figure 1:Dynamic of K using the JAYA approach.

Figure5: Result of states of example with input control

Figure 2: Estimates of the region of attraction of system

Figure6: Control input optimize.

Figure3: Result of states of example without JAYA.

You might also like