A Long-Range Adaptive Controller for Robot Manipulators

Abstract

The control of mechanical manipulators is an important application area for adaptive control. However, standard self-tuning controllers turn out to be nonrobust with respect to unmodeled dynamics. Attempting to overcome this limitation, long-range adaptive controllers are receiv ing an increasing interest. In this work an adaptive con trol structure based on a variant of the MUSMAR algo rithm is developed and tested on the control of a model for a two-link planar mechanical manipulator. Problems concerning the controller implementation are discussed.