Iterative Learning Control of a One-Link Flexible Manipulator

Abstract The paper describes the design and the implementation of a learning controller for the flexible forearm of a two link manipulator which is available in our Laboratory. Trajectory tracking based on iterative learning control is tested. In this control technique the required input for a given trajectory is built from successive experiments and the performances on the repetitive task are asymptotically improved from one iteration to the other. It works well even with a limited a priori knowledge of the system. Experimental results are shown in different working situations (planar trajectories with/without gravity effects) to demonstrate the control technique effectiveness.