Robust constrained state space ILC for 3DOF robot manipulator

Abstract

This paper focuses on the effect of the control system parameters on the convergence speed of two constrained state space Iterative Learning Control (ILC) algorithms: Bounder Error Algorithm (BEAILC) and Constrained Output Algorithm (COILC), applied to the nonlinear model of a 3DOF robotic manipulator in presence of recurring disturbance. Analysis and comparison of previously mentioned algorithms were conducted through simulations. The obtained results have shown that COILC algorithm converges faster than BEAILC algorithm, as the BEAILC restricts the output trajectory more rigorously. Simulations have shown that change in feedback parameters’ values has higher impact on the iteration interruptions (increase will lower the number of interruptions), while the learning parameters have higher impact on the whole ILC procedure duration (decrease will require more iterations to achieve the desired tracking accuracy). Additionally it’s been shown that both algorithms successfully rejected the recurring disturbance.