Adaptive control methods for mechanical manipulators: a comparative study.

Abstract

A new adaptive control law for mechanical manipulators that maintains uniformly good performance over a wide range of motions and payloads is developed. This control strategy combines properties from both the Model Reference Adaptive Control and the Self Tuning Regulator Theory and serves to extend the Adaptive Model Following Control approach into using a nonlinear reference model. The design procedure is simple resulting in an overall system which is globally stable and offers itself to microcomputer implementation. The effectiveness of the approach is demonstrated on several computer simulations which compares its performances against some of the commonly known adaptive control techniques. Also presented is a comparison of the computation complexity of different methods used in deriving the dynamic equations of motion of a mechanical manipulator as well as a survey of various robot control methodologies available in the literature today.