Model Based Iterative Learning Control Using Parameter Identification of Model Parameters

Abstract

In this paper, iterative learning control(ILC) method for industrial robot manipulators that repeat same operation is proposed. ILC method is reducing tracking error by using error of previous trial, so ILC doesn't require information of robots. Performance of ILC can be, however, improved by using the information.Parameter identification(ID) is used to estimate information of robots. When doing parameter ID, robot model is linearized and data of input torque and output angle when robot tracking operation trajectory is used. Performance of the algorithm was verified by simulating 1-link manipulator with joint flexibility.