Robust Fault Detection for Nonlinear Time-delay Systems Based on
Fuzzy Observer
TAO Hongfeng
1
, ZHANG Hua
1
, HU Shousong
2
1. Key Laboratory of Advanced Process Control for Light Industry Ministry of Education, Jiangnan University, Wuxi 214122
E-mail: taohongfeng@hotmail.com
2. College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, 210016,China
E-mail: hushousong@nuaa.edu.cn
Abstract: Robust fault detection method based on fuzzy sliding mode observers is developed for a class of uncertain nonlinear
systems with time-delay. The uncertain Takagi-Sugeno fuzzy model with time-delay is adopted for fuzzy modeling of nonlinear
system to reduce the modeling error, and a synthesis method of the design of sliding mode observers is introduced to generate the
estimate state error and residual then to eliminate the effect of the unknown uncertainties of the system for avoiding false alarms,
the synthesis method of the observers is presented via solving LMIs. Finally, a simulation example is given to show the design
procedure and illustrate the effectiveness of the proposed method.
Key Words: nonlinear time-delay systems, sliding mode observer, T-S fuzzy model, fault detection
1 Introduction
Uncertainties and time-delay are often encountered in
various engineering systems, such as biological, mechanical,
chemical systems and so on, which may lead to system
instability and degrade the performance of control process.
Yet the growing needs of reliability, efficiency and safety in
the presence of component and/or instrument fault of the
complex industrial system have made the process of reliable
control becomes an ever increasing important area of
research
[1-3]
.
It is well known that the observer technique is always a
fundamental, traditional, but important scheme for fault
detection and diagnosis
[4-8]
, and recently, various
approaches based on observers for detection have been
widely researched.
The Takagi-Sugeno(T-S) fuzzy models have been proven
to have great potential in application to complex nonlinear
systems using few rulers
[9-11]
, and the aggregation of all
subsystems represents the original nonlinear systems
adequately. Furthermore, the sliding mode controller is a
special variable structure controller in fact. Hence, the
concept of “sliding mode” can be introduced in the design of
observers to enhance robustness and reject disturbances and
also a class of uncertainties
[12]
. Considering these facts, we
can design a fuzzy sliding mode observer to estimate states
and generate residuals.
In this paper, we extend the results of robust
observer-based fault detection for a class of uncertain
nonlinear systems with time-delay, an effective robust fault
detection method is introduced based on the T-S fuzzy
model. In this work, in order to approximate the uncertain
nonlinear system, we have taken into account the time-delay
and uncertainty in the modeling of system to reduce the
*
This work is supported by National Natural Science Foundation of
China(61273070,61203092), Jiangsu Province Higher Education Natural
Science Basic Research(11KJB510007), 111 Project (B12018), and the
Priority Academic Program Development of Jiangsu Higher Education
Institutions
modeling error, then an sliding mode observer is introduced
to generate the estimate state error and residual and
eliminate the effect of the unknown uncertainties of the
system, the synthesis method of the observers is presented
via solving LMIs.
2 Problem Description
Consider a class of uncertain systems with time-delay and
unknown input as follows:
() ((),( ), ) ((),( ), )
((), )() ((), )() (,,)
() ( ())
xt f xt xt f xt xt
qxt ut pxt dt gxut
yt hxt
WT WT
TT
'
<
˄1˅
where
n
Rtx )( is the state vector ˈ
m
Rtu )( is the
control input vectorˈ
p
Rty )( is the output vectorˈ
s
Rtd )( denotes external disturbance vector,
is a
time-varying uncertain parameter;
)),(),((
W
txtxf ,
)),((
txq , )),((
txp , ))(( txh are known smooth nonlinear
functions;
)),(),((
W
txtxf is unknown uncertain
nonlinearities of system and modeling;
r
Rtuxg ),,( is
deemed to represent actuator fault vector. The matrix
rn
R
u
is constant.
W
is time delay, and fd
W
0 .
0)( tx ,
0t
.
Assumption The disturbance
)(td is
2
L -norm bounded
such as:
),()(
TU
xtd d , 0),( !
x .
Define the system ˄1˅can be approximated by a fuzzy
T-S model with time delay and unknown uncertainties of
system and modeling which is described by fuzzy If-Then
rules. The
i th rule of fuzzy model is in the following form:
i
˖IF )(
1
tz is
i
F
1
AND )(tz
v
is
v
F
1
THEN
Proceedings of the 33rd Chinese Control Conference
Jul
28-30, 2014, Nan
in
, China
3224