Neural Network Super-twisting based Repetitive Control for a Brushless DC Servo Motor with Parameter Uncertainty, Friction, and Backlash

Abstract

This paper presents a neural network super-twisting based repetitive control (NNSTRC) to improve the tracking accuracy of periodic signal. The proposed algorithm is robust against the plant uncertainty caused by the mass and viscous friction variation. Moreover, it compensates the nonlinear frictions, and the backlash by using the neural network based super-twisting algorithm. Firstly, a repetitive control (RC) is designed to track the periodic reference, and compensate the viscous frictions. Then, a stable neural network super twisting control (NNSTC) is constructed to compensate the nonlinear frictions, backlash, and plant uncertainty. The proposed algorithm is verified on a simulation model of rotational system. The simulation comparisons highlight the advantages of the proposed algorithm.