DESIGN AN ADAPTIVE ROBUST CONTROL USING NEURAL NETWORKS FOR CLEANING AND DETECTING ROBOT MANIPULATOR

Abstract

In this paper, an Adaptive Robust control using Neural Networks (ARNNs) is presented for Cleaning and Detecting Robot Manipulators (CLRM) in order to improve the positon tracking performance. To deal with the unknown dynamics of the CLRM, the ARNNs are applied in order to approximate the unknown dynamics. In addition, the robust sliding mode control (SMC) is used to eliminate the disturbances of the cleaning and detecting robot manipulator control system, compensate the estimation error. The online adaptive training laws of the controller are determined based on Lyapunov stability theorem. Therefore, the tracking performance, robustness and stability of the ARNNs for the CLRM are guaranteed. Moreover, the simulations performed on two-link cleaning and detecting robot manipulators are provided to prove the efficiency and robustness of the ARNNs.