Periodicity Locomotion Control Based on Central Pattern Generator

Abstract

In this article, a method to generate the periodical gait of a walking robot from the point of view of neurophysiologic is proposed. The neuraxon's nonlinear oscillation model, i.e. simplified Hodgkin-Huxley model is adopted. An octopod like robot's CPG (central pattern generator) consists of eight neuron oscillators, each of them is responsible to control the motion of an individual leg. The commands for oscillators are generated by high-level control system, which is responsible to generate some periodical motion modes and regulate switching between the modes by setting and tuning up parameters instead of controlling all details. The simulation results show that the designed central pattern generator model can generate V shape wave gait which satisfies the phasic demand, besides, the validity to control the periodical motion based on the central pattern generator is proved