Motion Planning for Snake Robots

Abstract

In this dissertation, we propose novel motion planning algorithms that solve the motion planning problem for planar snake robots locomoting in various environments: snakes floating in space, snakes swimming in viscous mediums, and kinematic snakes undulating on the ground. Motion planning for snake robots is usually considered challenging due to the actuation redundancy along the body of the snake robot and due to the various types of environmental forces acting on the snake robot.

In this dissertation, we exploit this redundancy to formulate the motion planning problem as an optimization problem. Not only does this formulation generate optimal shape trajectories, but also enables the problem to account for additional requirements such as obstacle avoidance, inter-link collision avoidance, and the avoidance of singular shape configurations.

We aim in this dissertation to propose motion planning algorithms applicable for snake robots of any dimension, that is, for snake robots having any number of links. The effectiveness and robustness of the proposed algorithms in generating shape trajectories is validated through dynamic simulations on snake robots of different dimensions. Additionally, the algorithm proposed for kinematic snake robots is experimentally validated on a planar kinematic snake robot prototype.