Dynamic Modeling and Fault-Tolerant Reconfiguration Strategies for PneumaticallyActuated Soft Snake Robot: Enabling Robust Locomotion in the Presence of Air Leakage

Abstract

Recently, the study of soft robots has become quite popular for their flexibility, safer interaction with humans, and low-cost adaptability to complex and uncertain environments. This paper presents a bond graph model of a pneumatically actuated soft snake robot and novel reconfiguration strategies for the soft snake robot to accommodate the faults caused by air leakage. With the proposed recon-figuration strategies, the soft snake robot can have robust locomotion along the desired trajectory even if an air leakage fault occurs during the locomotion. An efficient procedure of fault detection and isolation is also incorporated into the bond graph model of the robot. We conduct in-depth numerical experiments to validate the proposed robot model’s performance and reconfiguration strategies. Experimental results demonstrate that the soft snake robot, aided with reconfig-uration strategies, achieves straight-line locomotion with the lateral undulation gait despite air leakage faults.