Skip to Main Content
This paper presents an approach to the stability analysis of neural networks based adaptive controllers for motion control of robot manipulators. New adaptive feedback and feedforward control structures using neural networks are proposed. The controllers are adaptive to robot dynamics and payload uncertainties. Practical asymptotic stability conditions for the proposed controllers are given considering the neural networks learning errors. A robust adaptive approach which leads to global asymptotic stability is also presented. The analysis includes the evaluation of the control error as a function of the neural networks learning errors.