Perturbation wavelet neural sliding mode position control for a voice coil motor driver

Abstract

To cope with the nonlinear electro-magneto-mechanical characteristics, this paper proposes a perturbation wavelet neural sliding mode position control (PWSPC) system for a voice coil motor (VCM) driver. A perturbed wavelet neural network (PWNN) approximator is used to online approximate the unknown nonlinear term in the VCM system dynamics. The PWNN approximator uses perturbed wavelet functions to handle the rules uncertainties like interval type-2 fuzzy sets. The structure learning ability enables the PWNN approximator to evolve its structure online. Further, the parameter learning laws and stability analysis are derived in the sense of Lyapunov function; thus, the parameter convergence and system stability can be guaranteed. Finally, the experimental results verify that the proposed PWSPC system can achieve favorable control performance such as good disturbance rejection and good tracking accuracy.