Active Manipulation of Elastic Rods using Optimization-based Shape Perception and Sensorimotor Model Approximation

• Published in: 2022 41st Chinese Control Conference (CCC) pp. 3780 - 3785
• Main Authors: Ma, Guangfu, Qi, Jiaming, Lv, Yueyong, Zeng, Haibin
• Format: Conference Proceeding
• Language: English
• Published: Technical Committee on Control Theory, Chinese Association of Automation 25-07-2022
• Subjects: Adaptation models; Adaptive Kalman Filter; Deformable Objects; Feature extraction; Kalman filters; Robot motion; Robot sensing systems; Robotics; Sensor-based Control; Shape; Shape-servoing; Stability criteria
• ISSN: 2161-2927
• DOI: 10.23919/CCC55666.2022.9901658

Deformable object manipulation (DOM) remains challenging in robotics since the physical characteristics of deformable linear objects (DLO) are generally strongly nonlinear and unknown. This paper presents an active deformation frame-work for robot manipulating the elastic rod. The optimization-based curve fitting method is designed to extract a low-dimensional feature vector representing the three-dimensional centerline of the elastic rod. Deformation Jacobian Matrix (DJM) connecting feature changes and robot motion changes is estimated online by Adaptive Kalman Filter (AKF). The model-free velocity controller is designed based on the optimization criterion to accelerate the deformation speed of the elastic rod. The stability of the proposed manipulation framework and the boundedness of all signals in the closed-loop system are proved by Lyapunov theory. Detailed simulations are given to evaluate the performance of the proposed approach.