1 Introduction
Networked control systems (NCSs) are systems in which
control loops are closed over a communication network w-
hich may be share with other applications. Introducing net-
work into a control system brings many advantages over
point-to-point wired conventional control systems, includ-
ing lower cost, simpler installation and maintenance, and
better system flexibility. However, data exchange via net-
work inevitably results in transmission delays, transmission
intervals, packet losses, variable sampling and so on, which
will degrade system performance and even lead to closed-
loop instability [1], [2]. It is known that time-varying
packet losses and variable sampling can be equivalently
dealt with as time-varying transmission intervals. How to
compensate transmission delays and intervals in a linear
NCS is a fundamental problem.
In existing literatures, robust analysis and synthesis of li-
near NCSs have been extensively considered with respect
to time-varying transmission delays and intervals through
three main approaches, i.e. discrete-time model approach
[3]-[6], input-delay approach [7]-[15], and impulsive-mod-
el approach [16]-[20]. But relevant results are still very co-
nservative for uncertain NCSs with both time-varying tran-
smission delays and intervals. It is worth noting that distri-
bution of input delays is revealed to have an inherent rela-
tion with stability and performance of an NCS, and can be
exploited to reduce conservatism in the derived results
[21]-[25]. However, independent and identically distribu-
ted (i.i.d.) models of input delays are reasonable for NCSs
under a premise that the evolution of input delays lasts for a
sufficiently large number of transmission intervals. So the-
se models are often poor for an NCS in reality. Moreover,
This work was supported in part by the National Natural Science Foun-
dation of China under grants 61104105, 61374081, U0735003 and the
Natural Science Foundation of Guangdong Province under grant
S2013010013034.
average dwell-time approaches [24], [25] result in slow-sw-
itching constraints on the overall system while direct Lya-
punov-Krasovskii methods [21]-[23] may need further im-
provement to guarantee effectiveness of the obtained result-
s in the case of multiple distributed subintervals of input de-
lays.
It is well known that a nonlinear system can be
represented as a Takagi-Sugeno (T-S) fuzzy model which is
a weighted sum of some simple linear subsystems, and then
parallel distribution compensation scheme can be applied to
design controller with this model. Analysis and synthesis of
nonlinear NCSs based on T-S fuzzy models have received
increasing interest in the last decade [26]-[38]. Input-delay
approach was found to be widely adopted in existing
results, see e.g. [26]-[30] for uncertain, bounded and
fast-varying input delays, [31]-[33] for robust analysis
techniques, [34]-[36] for exploiting impulsive effects of
feedback signals and input delays as well as time derivative
of input delays, and [37] for i.i.d. random input delays. But,
to the best of our knowledge, it is still a real challenge to
obtain effective and less conservative performance
conditions based on T-S fuzzy model for a nonlinear NCS
with time-varying trans- mission delays and intervals.
Meanwhile, input missing is a common phenomenon in
control systems [24], [25], [39]. It may result from temporal
controller or actuator failure, intermittent unavailability of
controller, purposeful suspension of controller or actuator
to save energy, etc. When input missing occurs, the input of
plant is set to zero. Then control systems with input missing
are essentially switching systems. Input missing may cause
degradation of system performance and even instability of
the overall system. However, most existing researches on
NCSs with input missing are based on average dwell-time
approaches such that the derived performance conditions
additionally depend on average dwell time of subsystems
besides the rate of input missing [24], [25], [39]. It should
be mentioned here that few results are reported on fuzzy-
model-based stability of nonlinear NCSs with input
Fuzzy-Model-Based Robust Stability of Nonlinear Networked Control Systems
with Input Missing
Xiangyang Jia, Bin Tang, Defeng He, Shiguo Peng
School of Automation, Guangdong University of Technology, Guangzhou 510006, China
E-mail: jia_xiangyang@163.com, tangbin316@163.com, 13711524944@126.com, psg7202@126.com
Abstract: This paper is concerned with fuzzy-model based robust stability of nonlinear networked control systems
(NCSs) with time-varying transmission delays, transmission intervals and input missing based on a random-delay app-
roach. The real-time distribution of input delays resulting from transmission delays and intervals is modeled as a depen-
dent and nonidentically distributed process, and the occurrence of input missing is represented as a Bernoulli process.
Then a randomly switched Takagi-Sugeno fuzzy system with multiple input-delay subsystems is proposed to model the
nonlinear NCSs. Based on an improved Lyapunov-Krasovskii method, which takes into account the real-time distribu-
tion of input delays in estimating cross-product integral terms, new sufficient conditions are derived for the mean-square
robust exponential stability of the overall systems. Numerical examples are presented to substantiate the effectiveness of
our results.
Key Words: Networked control system, Takagi-Sugeno fuzzy model, Random-delay approach
1995
978-1-4799-3708-0/14/$31.00
c
2014 IEEE