Implementation of robust wavelet-neural-network sliding-mode control for induction servo motor drive

Abstract

This paper presents the design and properties of a robust wavelet-neural-network sliding-mode control (RWSC) system for an indirect field-oriented induction servo motor drive to track periodic commands. First, a total sliding-mode control (TSC) system with an integral-operation-switching surface, which is insensitive to uncertainties in the whole control process, is introduced. In the TSC system the controlled system has a total sliding motion without a reaching phase. Moreover, to relax the requirement for the bound of uncertainties, a wavelet-neural-network sliding-mode control (WSC) system is investigated to control the induction servo motor. In the WSC system, a wavelet neural network (WNN) is utilized to estimate the bound of uncertainties on line. In addition, an RWSC system is proposed to alleviate the chattering phenomena in the control effort. In the RWSC system, a boundary layer is introduced into the TSC system, and a WNN is used to estimate the width of the boundary layer and the uncertainty bound. Numerical simulation and experimental results due to periodic commands show that the dynamic behaviors of the proposed control systems are robust with regard to uncertainties. Furthermore, the advantages of the proposed RWSC system are indicated in comparison with the TSC and WSC systems.