Control Method for Robotic Manipulation of Heavy Industrial Cables

Fangli Mou,Dan Wu

DOI: https://doi.org/10.1109/ICMA52036.2021.9512586

2021-01-01

Abstract:In this paper, we address the challenge of robotic manipulation of heavy and stiff cables, the caused payload effect makes control degradation and can hardly be compensated. We propose a control framework for high precision manipulation that can greatly reject the payload effect caused by heavy cable objects. We model the cable payload effect by using the extracted cable features based on Mass-Spring-Damper method. This enables a real-time control methodology based on measurements of several feature points attached to the cable. The cable model serves as the prediction of payload effect and disturbances for our control scheme. The control method we proposed synthesizes the sliding mode control design with the active disturbance rejection control technique, an extended state observer is employed to estimate unknown disturbances and modeling errors which is hard to deal with in a conventional controller. We evaluated our method in simulation (with simulated sensor noise and experimental forces) on the 1.UR5 robot, our results reveal that our method can perform robot manipulation of heavy cables with rapidity and high precision (0.01 mm, same level as no-payload case).