Position Tracking of Parallel Robot by an Adaptive Second Order Sliding Mode Control based on Projection Neural Network

Abstract

A second-order sliding mode control method is utilized in conjunction with a projection neural network in this paper to steer the moving plate of the parallel robot. The proposed algorithm of this paper contains two programs as follows: first one is the optimal path generation program for the optimal platform movement, whereas the other one is the control torque generation program with adaptive second-order sliding mode control. The first program aims to move the optimal platform with the help of a projection neural network. In this program, the bounded performance index is introduced by considering the singularity constraint. Then, a dynamic model of a neural network is provided by converting it to a constrained optimization problem. Moreover, the second-order sliding mode control strategy is applied to remove the chattering phenomenon, which is based on the second-order derivative of the glide plane. Besides, an adaptive rule is employed to estimate the gain of instantaneous switching to improve the adaptability of the control algorithm.