A Neural Terminal Sliding Mode Control for tracking control of robotic manipulators in uncertain dynamical environments

Abstract

This article investigates a control algorithm for trajectory tracking control of robot manipulators in uncertain dynamical environments. To deal with chattering be-havior that always still exists in the conventional sliding mode control, to remove the requirement about a dynamic model of the uncertain robot system, and prior knowledge of the unknown function such as its upper boundary, the proposed solution is to develop a control method that combines the advantages of both ter-minal sliding mode control and radial basis function neural network. Further-more, the proposed controller’s fast stabilization and convergence are also signif-icantly improved by using a novel adaptive fast reaching control law. Hence, the proposed controller's performance expectations are always guaranteed such as high tracking accuracy, fast stabilization, chattering reduction, and fast conver-gence, robustness to uncertain dynamical environments. Especially, the proposed controller can operate without the robot's dynamic model. Both theoretical inves-tigations based on Lyapunov stability theory and computer simulation using MATLAB/Simulink are presented to confirm the effectiveness of the proposed control solution.