Learning control and its optimality: Analysis and its application to controlling industrial robots

Abstract

An adaptive control scheme based on iterative signal synthesis is called "learning control." Stability condition for a discrete linear time-varying dynamic system is analyzed, and the explicit condition for asymptotical stability of the system is obtained. Then an optimization schemes are applied to the proposed learning system and the learning control law is uniquely identified. We apply the obtained condition to an industrial robot and discuss its performance.