Hybrid adaptive fuzzy control of robot manipulators

Abstract

Detailed derivations of the Hybrid Adaptive Fuzzy Controller (HAFC) for robot manipulators are included in this thesis. This controller is a weighted combination of DAFC and IAFC together with a continuously-switched supervisory controller. The proposed controller is able to improve the performance of the system in an iterative manner by learning or adapting from the tracking errors that result from previous attempts at achieving the desired motion. In particular, the controller not only does not require the exact mathematical model of the manipulator, but also the boundary of the unknown dynamics can be adaptively estimated on-line.