A multirate sampling structure for adaptive robot control using a neurocompensator

Abstract

A novel multirate sampling structure is developed for adaptive control of robot manipulators. This control structure has the implementation advantage that the parameter adaptation in a control action is independent of the feedforward torque computation of the same control action. A fast sampling rate can be achieved by applying this structure. The parameter adaptation element in this structure is realized by a neurocompensator which is implemented using the ADALINE algorithm. Instead of the normal delta-learning rule used in ADALINE, a special learning rule is derived from the Lyapunov method to adjust the weights of the neurocompensator. Both system stability and error convergence can then be guaranteed. Simulation studies on a two-link manipulator show that the control system maintains very good trajectory tracking performance even in the presence of large parameter uncertainty and external disturbance. The satisfactory control performance of this approach is also demonstrated by experimental results for a one-link robot.