Control of robot manipulators in task-space under uncertainties using neural network

Abstract

In this paper, neural network-based controller is designed for the tracking control of robot manipulators in task-space under uncertainties. Especially, this controller does not need prior information of the upper bound of the unstructured uncertainties. By adaptively estimating the upper bound using feedforward neural network, effects of unstructured uncertainties can be eliminated and asymptotic error convergence can be obtained for the closed-loop system. Simulation results are carried out for a two-link elbow robot manipulator to show the effectiveness of the control scheme.