State feedback stabilization of singular systems with
both state and input delays
Zhou-Yang Liu, Chong Lin and Bing Chen
Institute of Complexity Science
Qingdao University, Qingdao 266071, P.R.China
Email: linchong
2004@hotmail.com
Abstract—This paper studies the stabilization problem for
a class of linear singular systems with both state and input
delays. Based on the neutral system approach, by employing
an augmented Lyapunov-Krasovskii functional, Wirtinger-based
inequality and some slack variables, sufficient conditions for the
design of state feedback controller are presented. The results are
established in the form of linear matrix inequalities, which can
be easily solved by Matlab software. Finally two examples are
provided to illustrate the effectiveness and the benefits of the
proposed method.
I. INTRODUCTION
Singular systems have extensive applications in economics,
circuits, chemical processes and other areas due to the fact that
a singular system model includes a set of coupled differential
and algebraic equations [1], [2]. Depending on the areas of
application, singular systems are also referred to as degener-
ate, descriptor, generalized, differential-algebraic or semi-state
systems [2]. In the past decades, many research papers have
been published to address the importance and the engineering
motivation of singular time-delay systems as time-delay is
often the main source of instability and poor performance
and such systems arise in a variety of physical systems such
as aircraft stabilization, chemical engineering systems, large-
scale electric network control and lossless transmission lines
(see [4]-[10] and the reference therein).
Various techniques have been proposed by many researchers
to deal with stability and stabilization of singular time-delay
systems. Among them, the Lyapunov-Krasovskii functional
(LKF) approach has been recognized as efficient and widely
used tool as it allows to derive efficient criteria which can be
recast into linear matrix inequalities (LMIs) and can be directly
solved by LMI tools in Matlab. Hence, for linear singular time-
delay systems, the last decade has seen a massive emergence of
research devoted to the construction of Lyapunov-Krasovskii
functionals. The problem of stability and stabilization has
been studied based on delay-independent method in [4], [5].
More recently, some new delay-dependent controller design
methods have been proposed in [6]-[9], [13], [14]. However,
it should be pointed out that all the aforementioned results
were obtained for systems with no input delays. In modern
industrial systems, sensors, controllers and plants are often
connected over a network medium and the sampling data
and thus controller signals are transmitted through a network
[11]. Therefore, network-induced delays and data dropout are
always inevitable. So it is important both in theory and practice
to consider the input delay when controller is designed.
This paper aims to design a controller to achieve the
asymptotic stability for a class of singular closed-loop systems
with both state and input delays, which has not been fully
investigated. The state and input delay is assumed to be
constant. Based on Lyapunov-Krasovskii functional approach,
Writinger-based inequality and neutral systems approach, a
state feedback controller is designed in terms of LMI. Several
numerical examples are included to illustrate the effectiveness
of the proposed method.
Notation: In this paper, R
n
denotes the n-dimensional real
Euclidean space; R
n×m
is the set of n × m real matrices;
diag{· · ·} denotes the block diagonal matrix, and the symbol
‘’ represents the symmetric elements in a symmetric matrix;
the superscript ‘T ’ denotes the matrix transportation; P>0
(P ≥ 0) means that P is a real, symmetric and positive definite
(positive semi-definite) matrix; I and 0 are, respectively, the
identity matrix and the zero matrix with compatible dimen-
sions.
II. P
ROBLEM FORMULATION AND PRELIMINARIES
Consider the following linear singular system with both
state and input delays:
E ˙x(t)=Ax(t)+A
d
x(t − τ
1
)+Bu(t − τ
2
), (1)
x(t)=φ(t),t∈ [−max(τ
1
,τ
2
), 0],
where x(t) ∈ R
n
is the state, u(t) ∈ R
m
is the control input.
E, A, A
d
∈ R
n×n
are known real constant matrices where
E may be singular, we shall assume that rank E = r ≤ n.
τ
1
> 0 and τ
2
> 0 stand for the constant time delay. φ(t) is
a continuously differentiable vector-valued initial function.
The control input u(t) is described by u(t)=Kx(t), where
K is the feedback gain to be designed. Applying the controller
to system (1), we have
E ˙x(t)=Ax(t)+A
d
x(t − τ
1
)+B
k
x(t − τ
2
), (2)
where B
k
= BK.
Inspired by the neutral system approach in [10], we try
to transform system (1) into a neutral form under certain
constraint. To proceed, the following definitions and lemmas
are needed.
Definition 1[2]. System (1) is said to be regular if the charac-
teristic polynomial, det(sE−A) is not identically zero; System
1593
978-1-4799-8389-6/15/$31.00
c
2015 IEEE