A UNIFIED ADAPTIVE LEARNING CONTROLLER FOR ITERATIVE AND REPETITIVE LEARNING OF ROBOTIC SYSTEMS

摘要

This paper presents a unified learning controller which learns the entire span of periodic and nonperiodic robot trajectories within a finite time interval. The learning controller learns the uncertain parameters of robot systems as well as the desired control inputs including the unknown external disturbances. It is shown that all the learning signals in the learning system are bounded and the robot motion converges to the desired one as the learning continues.