A robust hybrid controller for robotic manipulators using neural and H/sub /spl infin// techniques

A robust hybrid control method that integrates a neural controller, and a H/sub /spl infin// controller is proposed. The neural controller is designed for feedback linearisation and the H/sub /spl infin// controller is used to minimize the effect of parameter uncertainties of the model on the tracking performance. Simulations are provided to show that the disturbances generated from the model error will be quickly compensated and hence more robust performance can be achieved.