Dynamic modeling and analysis for planar peristaltic locomotion of a metameric earthworm-like robot

Abstract

The metameric earthworm-like robot can exhibit a wide variety of locomotion gaits across diverse environments. However, previous analyses of its planar peristaltic locomotion have largely relied on kinematic models, which assume ideal anchoring and actuation while overlooking practical challenges like limited propulsion and insufficient friction. This has led to significant discrepancies between kinematic predictions and experimental results. In response, this study introduces a comprehensive dynamic model for the planar peristaltic locomotion of a metameric earthworm-like robot. The model incorporates actuation and friction forces, which are absent in kinematic models, to more accurately predict the robot’s locomotion performance and trajectory. A proportional-derivative (PD) control method characterizes the actuation force and torque, ensuring limited actuation while achieving specified configurations. A modified anisotropic Coulomb’s dry friction model describes the resistive interaction between the bristles and the ground. An amplification factor captures the friction amplification due to radial expansion of robot segments, and a correction accounts for the lateral friction of the robot head segment, which lifts significantly during locomotion. The friction parameters are determined through extensive experimentation. Comparative locomotion tests across various gaits reveal that the dynamic model significantly improves the accuracy of predicting locomotion metrics and trajectories, addressing the limitations of kinematic models. The dynamic model aligns closely with experimental results, particularly in sidewinding and circular gaits, and effectively captures non-smooth stick-slip dynamics and insufficient forward propulsion. These improvements highlight the dynamic model’s superior predictive accuracy for earthworm-like robot locomotion, providing a foundation for future advancements in robotic design, analysis, and control.