Adaptive Tracking Control for a Class of Unknown Nonlinear
Time-Delay Systems Using Nonlinearly Parameterized Fuzzy
Approximators
Ping Li
1
, Xiaochao Zhang
1
, Fujiang Jin
1
1. College of Information Science and Engineering, Huaqiao University, Xiamen, Fujian, China, 361021
E-mail: pingping
−
1213@126.com
Abstract: A class of strict-feedback nonlinear systems with unknown system functions and time delays are considered in
this paper. Nonlinearly parameterized fuzzy systems with adaptive mechanism are employed to approximate the unknown
functions in a backstepping procedure. By using Taylor series expansion, the parameters are separated successfully, then
adaptive laws can be designed to update them on-line. The advantage of such a design is that the basis functions of
the fuzzy approximators are not require known before design. The proposed controller can guarantee the closed-loop
stability and good tracking performance.
Key Words: Adaptive, Fuzzy approximation, Backstepping, Nonlinear systems, Time-delay
1 INTRODUCTION
At present, effective design of nonlinear systems is still
a challenge to researchers, especially for those with un-
known system functions. In early years, Wang proved that
fuzzy logic systems can approximate real continuous non-
linear functions to any precision on given compacts, and
proposed a stable adaptive fuzzy controller in [1]. The
method was further developed in [2] and [3] for SISO and
MIMO systems, respectively. However, these results are
obtained on the basis that the controlled systems must sat-
isfy matching conditions. In order to control unknown non-
linear systems with mismatched conditions, adaptive fuzzy
approximation and backstepping design were combined in
[4, 5, 6]. Later, Chen proposed novel adaptive fuzzy control
schemes which contain less adaptive parameters in [7] and
[8]. Though the methods mentioned above achieved de-
sired performances, they are obtained with the restriction
that the fuzzy basis functions must known before design.
When the basis functions of the fuzzy approximators are
unknown, they may failed. Recently, Li proposed adaptive
fuzzy control schemes in [9, 10] using nonlinearly param-
eterized approximators. However, these results are limited
to delay-free systems.
Time delays are frequently encountered in practice engi-
neering systems. The existence of time delays may destroy
the stability and degrade the performance of the controller
system. The results on stabilizing nonlinear time-delay sys-
tems with approximation-based control have been devel-
oped in [11, 12, 13, 14]. Based on these results, the authors
This work is supported by National Nature Science Founda-
tion(61143005, 61273069), Technology Programme Significant Project of
Fujian Province(2011H6019), Technology Programme Important Project
of Fujian Province(2009H0033), Nature Science Foundation Youth Inno-
vation Project of Fujian Province(2011J05153), Technology Programme
Significant Project of Quanzhou City2008ZD14-21).
of [15, 16] presented adaptive fuzzy control methods for
output tracking of nonlinear time-delay systems, in which
the total number of adaptive parameters were reduced con-
siderably. However, the mentioned results are based on lin-
early parameterized fuzzy approximators, that is the fuzzy
basis functions of the approximators must be known. When
there are no much a priori knowledge to determine the basis
functions, they are not applicable.
Based on the above observation, this paper is concerned
with the adaptive control of unknown nonlinear time-delay
systems with nonlinearly parameterized fuzzy approxima-
tors. Inspired by the work in [9, 10], the control law is de-
rived from a backstepping procedure. And in each step, a
nonlinearly parameterized fuzzy approximator is employed
to approximate the packed unknown function. The main
difficulty encountered in the control design is that how to
design the adaptive laws for parameters with nonlinear re-
lationship. To overcome this difficulty, Taylor series expan-
sion is employed to separate the nonlinear parameters, then
adaptive laws can be designed for them. The proposed con-
troller guarantees that all signals in the closed-loop system
are bounded, besides, the output tracking error converges to
an arbitrarily small neighborhood of the origin. Compared
with the existing results, the method in this paper is less
dependent on a priori knowledge, which makes the control
scheme more applicable for practice engineering.
The rest of this paper is organized as follows. The prob-
lem statement and preliminaries are presented in Section II.
Then, the control design with the analysis is given in Sec-
tion III. In Section IV, a simulation example is employed
to demonstrate the effectiveness of the proposed control
scheme. Finally, Section V concludes the paper.
31
978-1-4673-5534-6/13/$31.00
c
2013 IEEE