Trajectory Planning on Rolling Locomotion of Spherical Movable Tensegrity Robots with Multi-Gait Patterns.

Abstract

Spherical movable tensegrity robots, resorting to the intrinsic hallmark of being lightweight and resilient, have exhibited tremendous potential in exploring unpredictable terrains and extreme environments where traditional robots often struggle. The geometry of spherical tensegrities is suitable for rolling locomotion, which guarantees the system to react to changing demands, navigate unexplored terrain, and perform missions even after suffering massive damage. The objective of this article is to enrich the type of spherical movable tensegrity robots with multiple kinematic gait patterns and to gain superior motion paths that are in conformity with the intrinsic features of structural rolling locomotion. Aiming at this purpose, three 12-rod spherical tensegrities with multi-gait patterns are investigated, and the dynamic simulation on independent (or evolutionary) gait patterns is conducted and testified on ADAMS. The routing spaces and the blind zones formed by single kinematic gait are compared to assess the suitability of the assigned kinematic gait pattern. Accordingly, we develop a trajectory planning method with the embedding of the steering control strategy into a modified rapidly exploring random tree (MRRT) algorithm to produce qualified marching routes. In the meantime, two momentous evaluation indictors, applicable to multi-gaits tensegrities, are introduced in searching the corresponding optimal gait patterns that conform to specified needs. The techniques are illustrated and validated in simulation with comparisons on several prototypes of tensegrity robots, indicating that the proposed method is a viable means of attaining marching routes on rolling locomotion of spherical movable tensegrity robots.