Adaptive Asymptotic Control for a Class of Uncertain Nonlinear Systems

Hanqiao Huang,Renwei Zuo,Zongcheng Liu,Shuangyu Dong

doi:10.1109/access.2019.2926264

Abstract

This paper addresses the asymptotic tracking problem of adaptive neural control for a class of uncertain strict-feedback nonlinear systems. As a universal approximator, the neural network is widely utilized to solve the tracking control problem of unknown continuous nonlinear systems. Due to the existence of neural network approximation errors, previous neural network-based control approaches can only achieve the bounded tracking rather than the asymptotic tracking. This paper designs an asymptotic error eliminating term to achieve the adaptive neural asymptotic tracking. By utilizing the Lyapunov stability theory, all the variables of the resulting closed-loop system are proven to be semi-globally uniformly ultimately bounded, and the tracking error can converge to zero asymptotically by choosing design parameters appropriately. A simulation example is presented to show the effectiveness of the proposed control approach.

Highlights

Over the past few decades, adaptive control for a class of strict-feedback nonlinear systems with parameterized functions or matched uncertainties have been extensively studied for both theoretical interests and engineering applications [1]–[5]
The early stages of the research cannot always be applied because some practical systems inevitably contain some unknown functions which cannot be expressed as the linearized parameter form, and the unknown uncertainties may not appear in the same channel as the control input
(1) In this paper, we develop an adaptive neural-networkbased asymptotic tracking controllers for a class of uncertain strict-feedback nonlinear systems

Summary

INTRODUCTION

Over the past few decades, adaptive control for a class of strict-feedback nonlinear systems with parameterized functions or matched uncertainties have been extensively studied for both theoretical interests and engineering applications [1]–[5]. To solve the controller design problem of nonlinear systems with unknown functions and mismatched uncertainties, many researchers resorted to the backstepping technique and neural network [6]–[8]. In [37], with the aid of barrier functions, a universal adaptive state-feedback asymptotic tracking control strategy is proposed for a class of unknown time-varying nonlinear systems. Motivated by the above discussion, in this paper, an adaptive neural control scheme is proposed for a class of uncertain strict-feedback nonlinear systems in the frame of backstepping method. (1) In this paper, we develop an adaptive neural-networkbased asymptotic tracking controllers for a class of uncertain strict-feedback nonlinear systems.

PROBLEM STATEMENT AND PRELIMINARIES

RBFNN BASICS

STABILITY ANALYSIS

SIMULATION RESULTSION

Findings

CONCLUSION