Bipedal locomotion modeled as the central pattern generator (CPG) and regulated by self organizing map for model of cortex

This paper describes a biologically inspired algorithm mimicking locomotion, and associated brain inspired software architecture to model bipedal gait. The central pattern generator (CPG) neural network is first modeled and next, self-organizing maps of gait for running and walking are created. This biological or brain inspired neural network model is finally assimilated to account for cortical control or modulation of gait. This work demonstrates the utility of using the CPG neural networks and self-organizing maps as the controller to mimic normal and abnormal gait patterns. The work presented here is our first step towards developing a biologically inspired neuroprosthetic device to help stroke patients regain normal gait at a faster pace.