This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
IEEE TRANSACTIONS ON CYBERNETICS 1
Decentralized Adaptive Event-Triggered H
∞
Filtering for a Class of Networked Nonlinear
Interconnected Systems
Zhou Gu ,PengShi , Fellow, IEEE, Dong Yue, Senior Member, IEEE,
and Zhengtao Ding
, Senior Member, IEEE
Abstract—This paper focuses on the issue of designing an
adaptive event-triggered scheme to the decentralized filtering for
a class of networked nonlinear interconnected system. A novel
adaptive event-triggered condition is proposed by constructing
an adaptive law for the threshold. This new type of threshold
mainly depends on the error between the states at the current
sampling instant and the latest releasing instant, by which the
data release rate is adapted to the variation of the system. The
limitation of network bandwidth is alleviated on account of a
large amount of “unnecessary” packets being dropped out before
accessing the network. Sufficient conditions are derived such that
the overall filtering error system under the proposed adaptive
data-transmitting scheme is asymptotically stable with a pre-
scribed disturbance attenuation level. An example is given to
show the effectiveness of the proposed scheme.
Index Terms—Adaptive event-triggered scheme, filtering, non-
linear networked interconnected system, Takagi–Sugeno (T–S)
fuzzy model.
I. INTRODUCTION
L
ARGE-SCALE system has gained a growing attention
since the 1970s [1], [2]. It has many practical applica-
tions, such as transportation systems [3], power systems [4],
multiagent systems [5], aerospace vehicles [6], network of
Chua’s chaotic circuits [7], and ecosystems [8]. Such a
Manuscript received July 28, 2016; revised October 16, 2017; accepted
January 22, 2018. This work was supported in part by the National Natural
Science Foundation of China under Grant 61473156, Grant 61533010, Grant
61773131, and Grant U150921, in part by the 111 Project under Grant B17048
and Grant B17017, and in part by the Australian Research Council under Grant
DP170102644. This paper was recommended by Associate Editor W. Wang.
(Corresponding author: Zhengtao Ding.)
Z. Gu is with the College of Mechanical and Electronic Engineering,
Nanjing Forestry University, Nanjing 210037, China (e-mail:
gzh1808@163.com).
P. Shi is with the School of Electrical and Electronic Engineering,
University of Adelaide, Adelaide, SA 5005, Australia, and also with the
College of Engineering and Science, Victoria University, Melbourne, VIC
8001, Australia (e-mail: peng.shi@adelaide.edu.au).
D. Yue is with the Institute of Advanced Technology, Nanjing
University of Posts and Telecommunications, Nanjing 210023, China (e-mail:
medongy@vip.163.com).
Z. Ding is with the School of Electrical and Electronic Engineering,
University of Manchester, Manchester M13 9PL, U.K. (e-mail:
zhengtao.ding@manchester.ac.uk).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TCYB.2018.2802044
system is composed of several connected low-order subsys-
tems, which is also called as an interconnected system. A
common feature of such a system is that the subsystems
or components are usually widely located in space [9]–[12].
Decentralized strategy allows the control implementation to be
more feasible and flexible. But signal transmission becomes
more complicated when using point-to-point wired connec-
tion under this control strategy. Wireless/wired communication
network is an alternative to transform the control system
into a networked control system (NCS). The information is
exchanged over the network among subsystems and control
components. However, the limitation of network bandwidth
or some other factors restraining the network, such as, the
consumption of power while using wireless network, will
lead to increasing difficulties and challenges of analysis and
synthesis [10], [13].
Nonlinearity is an inherent nature of many practical systems
that should not be neglected in modeling system. In the past
decades, research on control/estimation problems for non-
linear systems has received considerable attention [14]–[17].
Takagi–Sugeno (T–S) fuzzy model technique has been proved
to be a successful approach in dealing with the prob-
lem of modeling and analyzing nonlinear dynamic control
systems [18]–[20]. Based on T–S fuzzy mode, an H
∞
fil-
ter design for a class of continuous-time/NCSs with multiple
state-delays was investigated in [14]. The objective of H
∞
fil-
tering is to design an estimator for a given system such that
the L
2
gain from the exogenous disturbance to the estimation
error is less than a given level [21]. Xu et al. [22] investigated
the H
∞
filtering for discrete-time nonlinear systems using T–S
fuzzy model with consideration of multiple sensor faults. The
sensor saturation and missing measurements were considered
in the filtering of T–S fuzzy NCSs in [23] and [24]. For NCSs,
there are three ways to improve the performance of the control
system.
1) By improving the method of control design to adapt
network inherent defectives, such as packet drop-out,
network-induced delay, etc.
2) By enhancing the network quality of service (QoS),
such as the method of quantization, event-triggered
mechanism, etc.
3) By co-designing method.
Zhang et al. [25] studied the problem of quantized H
∞
filtering
for T–S fuzzy systems. However, quantization method is not
2168-2267
c
2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.