A Nonlinear Compensator Using Neural Networks for Robotic Manipulators

Abstract

This paper presents a nonlinear compensator using neural networks for trajectory control of robotic manipulators. The nonlinear compensator has a new architecture using both the computed torque method with the model of the manipulator and the neural networks. The neural networks in the compensator are to compensate only the parameter deviations and unmodeled effects of the manipulator and do not have to compensate all the nonlinearities of the manipulator. The neural networks are efficient in learning.