FEM-based Trajectory Tracking Control of a Soft Trunk Robot

Ke Wu,Gang Zheng,Junfeng Zhang

DOI: https://doi.org/10.1016/j.robot.2021.103961

IF: 3.7

2022-01-01

Robotics and Autonomous Systems

Abstract:As a novel class of robots, soft robots have demonstrated many desirable mechanical properties than traditional rigid robots due to their nature of being compliant, flexible and hyper-redundant, such as great adaptability to unknown environments, safe human robot interaction (HRI), energy-saving actuation and the maneuverability to display diverse mechanical properties. However, its inherent high-DoF nature would result in some complex nonlinear behaviors, and their kinematic or dynamic models are therefore harder to deduce than the ones of conventional rigid robots. In this paper, we propose a trajectory tracking control strategy for a soft trunk robot based on Finite Element Method (FEM). We first plan a feasible trajectory for the studied robot in SOFA (a FEM-based simulator) by solving a model-prediction-control (MPC)-based optimization problem. The second step is to conduct linearization around the pre-designed trajectory, based on which an associated controller can be then developed. The detailed derivation of the mentioned work is explained accordingly. In the end, the results of experimental validation is presented to prove the feasibility of the proposed method.