Gaussian neural network for direct adaptive control of robotic systems

Abstract

We treat the question of direct adaptive control of multi-link robotic systems with revolute joints. It is assumed that the dynamics of the robotic systems are not known. A radial basis function (RBF) neural network is used in a feedback loop for the control. An adaptive law is derived for the joint angle trajectory tracking. Simulation results are presented to show that in the closed-loop system precise trajectory control is accomplished. Furthermore, the effect of choice of number of neurons in the RBF neural network on the performance of the controller is also examined. >