Spatial locomotion of a metameric earthworm-like robot: generation and analysis of gaits.

This research aims to advance the current state-of-the-art of earthworm-like robots by extending the gait study from rectilinear and planar locomotion to spatial locomotion. In detail, the following five important but largely unaddressed issues are tackled in this paper, including the multibody kinematic modeling of the robot with spatial locomotion capability, 3D gait generation algorithm, spatial locomotion mode classification, spatial locomotion performance evaluation, and 3D gait planning. Based on the multibody model of the metameric earthworm-like robot and learning from the retrograde peristalsis wave, a gait-generation algorithm is proposed for generating all admissible 3D locomotion gaits. The 3D locomotion can be classified into four major types (rectilinear, sidewinding, circular, and cycloid), and further, ten modes with their defining conditions in terms of the gait parameters being derived. Among them, several spatial locomotion modes, such as the 3D cycloid locomotion, are uncovered for the first time. For each mode, different kinematic indexes, including the average (axial) velocity and the circumradius of the trajectory, are evaluated to comprehensively clarify the relationship between the locomotion performance and the gait parameters. Moreover, aiming at reaching the target both quickly and precisely, 3D gait planning is also studied. With contradictory objectives, the Pareto front is determined via the genetic algorithm so that a trade-off between the proximity and speed to the target can be achieved. [ABSTRACT FROM AUTHOR]