Iterative control and its application to motion control of robot arm - A direct approach to servo-problems

Abstract

There is a new and direct servo-control method called "Learning Control" which has a good advantage that the command input is not necessary a function generated from time-invariant linear systems but is required in conventional servo-theory using "Internal Model Principle". Therefore, this new theory is very attractive since it may estabilish a new foundation of servo-theory in which the shape of the command is arbitrary. The idea is very simple, we determine the input function U(s) satisfying R(s)=G(s)U(s) using some iterative method. In this paper, we derive conditions on the convergence of this method and practical design procedures, and the results are applied to the trajectory control of robot arm such that the tip of the arm draws any shape with satisfactory precision in several trials.