Central pattern generator control of a differential wheeled robot

Bio-inspired central pattern generator approaches have been used for robotic autonomous locomotion mainly in legged, serpentine and worm like robots. In this paper we propose a new bio-inspired central pattern generator (CPG) neural circuit to control a differential wheeled robot. The CPG uses bursting neurons coupled through time evolving synapses, which endows them with rich rhythm negotiation capabilities. The overall circuit activity produces robust yet flexible rhythms. We show that this CPG can produce sequential command signals to rhythmically control the motion of the differential wheeled robot in a situation where the wheels can only partially rotate.