Adaptive Fuzzy Sliding Mode Control Design : Lyapunov
Approach
H.F. Ho, Y.K. Wong and A.B. Rad
Department of Electrical Engineering,
The Hong Kong Polytechnic University,
Hong Kong
e-mail:
{hfho.ee,eeykwong,eeabrad}@polyu.edu.hk
Abstract
An adaptive fuzzy sliding mode control algorithm is
proposed for a class of continuous time unknown nonlinear
systems. In contrast to the existing sliding mode control
(SMC) design, where the presence of hitting control may
introduce problems to controlled systems, the proposed
adaptive fuzzy logic controller takes advantages of both
SMC and proportional integral (PI) control. The chattering
action is attenuated and robust performance can be ensured.
The stability analysis for the proposed control algorithm is
provided. Two nonlinear system simulation examples are
presented to verify the effectiveness of the proposed
method.
1 Introduction
Fuzzy logic control is a technique of incorporating expert
knowledge in designing a controller. Past research of
universal approximation theorem [1] shown that any
nonlinear function over a compact set with arbitrary
accuracy can be approximated by a fuzzy system. There
have been significant research efforts on adaptive fuzzy
control for nonlinear systems [2]-[4].
It is well known that the sliding mode control method
provides a robust controller for nonlinear dynamic systems
[5], [6]. However, it inherits a discontinuous control action
and hence chattering phenomena will take place when the
system operates near the sliding surface. One of the
common solutions for eliminating this chattering effect is to
introduce a boundary layer neighboring the sliding surface
[6], [7]. This method can lead to stable closed loop system
without the chattering problem, but there exists a finite
steady state error due to the finite steady state gain of the
control algorithm.
The adaptive fuzzy controller incorporating the fuzzy logic
and the sliding mode control (SMC) [8]-[10] for ensuring
stability and consistent performance is an active research
topic of the fuzzy control. In particular, this research work
integrates the fuzzy approximation theory and the SMC into
the fuzzy logic controller. These approaches are similar in
the aspect of directly approximating the sliding mode
control law by fuzzy approximators. One of the
advantages
of this control strategy is insensitive to modeling
uncertainty and external disturbances. Many adaptive fuzzy
sliding mode control (AFSMC) schemes have been
proposed and the chattering phenomena in the controlled
system can be avoided by using the fuzzy sliding surface in
the reaching condition of the SMC [11-13]. However, these
features make the number of fuzzy rules increasing with the
complexity of the fuzzy sliding surface involved. As the
sliding mode control law can separated into two parts i.e.
the equivalent control and the switching control [10]. The
role of the controller is to schedule these two components
under different operating conditions. In order to improve
the steady state performance of the AFSMC, an adaptive
fuzzy logic controller combining a proportional plus
integral (PI) controller and the SMC is considered in this
paper. The proposed control scheme provides good transient
and robust performance. Moreover, as the proposed
controller integrates the PI control with the SMC, the
chattering phenomenon can be avoided. In this paper, it is
proved that the closed-loop system is globally stable in the
Lyapunov sense and the system output can track the desired
reference output asymptotically with modeling uncertainties
and disturbances.
This paper is organized as follows. First, the problem
formulation is presented in Section 2. A brief description of
fuzzy logic system is included in Section 3. In section 4, the
adaptive fuzzy sliding control is proposed. Simulation
results for the proposed control concept are shown in
Section 5. Finally, the paper is concluded in Section 6.
2 Problem Statement
Consider a general class of SISO n-th order nonlinear
systems as follow form [5]:
xy
tdutxgtxfx
n
)(),(),(
)(
(1)
where
f and
are unknown nonlinear functions,
nT
n
Tn
Rxxxxxxx
],,,[],,,[
21
)1(
is the state
vector of the systems which is assumed to be available for
measurement,
Ru
and
Ry
are the input and the