Bio-inspired Locomotion Control for UBot Self-reconfigurable Modular Robot

摘要

This paper first presents a mathematic CPG (central pattern generator) model which has been developed based on the characteristics of a self-reconfigurable modular robot (UBot)'s modules with universal joints. Then, a bionic motion neural control network based on the CPG is proposed to solve the problem of multi-mode locomotion control problem in the complex environment. The bionic network is composed of perceptual neurons, CPG phase modulation network and motor neurons, so it can coordinate the walking and creeping gait of the modular robot before and after deformation, and adapt to autonomous movement in the complex environment with challenging features, such as steps, slopes and obstacles. Finally, the proposed motion control algorithm is verified by experiments.