A General Kinematic Model for Multimodal Locomotion in Bioinspired Robots.

Locomotion in animals such as fish, snakes, inchworms, and octopuses exhibits a remarkable diversity, with each species utilizing distinct body morphologies and movement strategies. Currently, no existing kinematic model is capable of describing the full range of locomotion exhibited by these animals. Addressing this challenge holds important implications for both the study of biomechanics of animals and the development of bioinspired robots. In this work, we propose a general kinematic model that integrates the curvature equation with a nonlinear oscillator. Through parameter adjustments, its morphology can transition between the motions of various animals. It is the most versatile kinematic model to date for describing multimodal locomotion of animals so far as we know. By translating the general kinematic model into a motion control algorithm and combining it with virtual simulation, we create a motion optimization framework that substantially simplifies the complexity of multimodal control for bionic robots with diverse actuation mechanisms, thereby enhancing their maneuverability. Using fish locomotion as an example, we validate the methodology on an untethered multijoint robotic fish, successfully enabling the robotic fish to perform cruising and various fast turn motions, thereby demonstrating its effectiveness in guiding motion control. This work is believed to have laid the foundation for the study of bionic motion and bioinspired robots.