Journal of Systems Engineering and Electronics Vol. 21, No. 4, August 2010, pp.650–657
Available online at www.jseepub.com
Robust stability and stabilization for uncertain discrete-time
switched singular systems with time-varying delays
Jinxing Lin
1,∗
, Shumin Fei
1
, and Jiong Shen
2
1. Key Laboratory of Measurement and Control of CSE, Ministry of Education, School of Automation,
Southeast University, Nanjing 210096 , P. R. China;
2. School of Energy and Environment, Southeast University, Nanjing 210096, P. R. China
Abstract:
The problems of robust stability and stabilization via
memoryless state feedback for a class of discrete-time switched
singular systems with time-varying delays and linear fractional
uncertainties are investigated. By constructing a novel switched
Lyapunov–Krasovskii functional, a delay-dependent criterion for
the unforced system to be regular, causal and uniformly asymp-
totically stable is established in terms of linear matrix inequalities
(LMIs). An explicit expression for the desired memoryless state
feedback stabilization controller is also given. The merits of the
proposed criteria lie in their less conservativeness and relative sim-
plicity, which are achieved by considering additionally useful terms
(ignored in previous methods) when estimating the upper bound
of the forward difference of the Lyapunov–Krasovskii functional
and by avoiding utilizing any model augmentation transformation.
Some numerical examples are provided to illustrate the validity of
the proposed methods.
Keyw ords: discrete-time switched singular system, time-varying
delay, delay-dependent criterion, switched Lyapunov–Krasovskii
functional, robust stability, stabilization.
DOI: 10.3969/j.issn.1004-4132.2010.04.00
1. Introduction
Switched systems belong to a special class of hybrid con-
trol systems, which consist of a family of continuous-
and/or discrete-time subsystems, and a switching rule
specifying the switching among them. Recently, switched
systems h ave received increasing attention because the
study for these systems is not only academically challeng-
ing, but also of practical importance [1–3].
Singular time-delay systems, which are also referred
to as descriptor time-delay system s or implicit time-delay
systems, often appear in various engineering systems, for
Manuscript received December 26, 2008.
*Corresponding author.
This work was supported by the National Natural Science Foundation
of China (60904020; 60835001) and the Jiangsu Planned Projects for
Postdoctoral Research Funds (0802010C).
example, aircraft attitude control, flexible arm control of
robots, large-scale electric network control, chemical en-
gineering systems, lossless transmission lines [4,5]. In the
past decades, control for singular time-delay systems has
been extensively studied and many notions and results in
regular time-delay systems have been extended to singular
time-delay systems [6–11]. For switched singular time-
delay systems (SSTDs), since regularity, absence of im-
pulses (for continuous systems) or causality (for discrete
systems), state consistence and stability should be consid-
ered at the same time, the analysis and synthesis of such
systems are much more complicated [12,13] and, to date,
the corresponding results are relatively few [13–16]. In
the continuous setting, the robust stability problem was
investigated in [14] by using multiple Lyapunov function
technique s. In [15], the stability and stabilization prob-
lems for Markovian switching singular time-delay systems
were discussed based on delay-independent LMI. For un-
certain discrete-time switched singular systems with con-
stant time-delay, Ma et al. [13] solved the delay-dependent
robust stability, stabilization and H
∞
control problems
by tran sforming the SSTDs into switched regular time-
delay systems. Based on the model augmentation trans-
formation, Ma et al. [16] also studied the robust stabil-
ity and H
∞
control problems for discrete-time Markovian
switching singular systems with mode-dependent time-
delay. However, there is much room for further inves-
tigation of discrete-time SSTDs. Firstly, the model aug-
mentation transformation method used in [13,16] is not al-
ways implementable as the dimension of the augmented
system increases with that of the original system. That
is, when the dimension of the SSTDs is large, the aug-
mented system will become much complex, and thus is
difficult to analyze and synthesize. Secondly, in [16], the
calculation of ΔV (k) led to −
k−1
l=k−
¯
d+1
y
T
l
Z
1
y
l
.Inorder