Stable Sampled-data Adaptive Control of Robot Arms Using Neural Networks

Abstract

Stable neural network-based sampled-data indirect and direct adaptive control approaches, which are the integration of a neural network (NN) approach and the adaptive implementation of the discrete variable structure control, are developed in this paper for the trajectory tracking control of a robot arm with unknown nonlinear dynamics. The robot arm is assumed to have an upper and lower bound of its inertia matrix norm and its states are available for measurement. The discrete variable structure control serves two purposes, i.e., one is to force the system states to be within the state region in which neural networks are used when the system goes out of neural control; and the other is to improve the tracking performance within the NN approximation region. Main theory results for designing stable neural network-based sampled data indirect and direct adaptive controllers are given, and the extension of the proposed control approaches to the composite adaptive control of a flexible-link robot is discussed. Finally, the effectiveness of the proposed control approaches is illustrated through simulation studies.