THREE-LAYER CPG MODEL FOR BIPED ROBOT

Bin He and Qiang Lu

Keywords

CPG, biped robot, human locomotion

Abstract

The central pattern generator (CPG) has been identified as a controller responsible for locomotion movements in spinal cords. In this paper, the experimental data are collected from joints’ positions of 10 healthy adults as they are walking for 3 min at various speeds on a treadmill. The experimental results show that the speed changes the frequency in locomotion and the rhythm pattern consists of the main rhythmic signal and the minor pattern information. To explain the phenomenon, it is supposed that there exists a neuron which is related to the change in the speed in locomotion and a rhythm- generation model with three neurons which include the extensor neuron, the flexor neuron and the speed neuron is established. The stability of this model is analysed and the conditions for stable oscillation are then obtained. Consistent with neurophysiological studies, a new CPG model is based on three layers which include the rhythm pattern, the activation and the motoneuron layer. As an illustration, this new CPG model is applied in robot gait. In conclusion, the new three neurons rhythm-generation model built up in this paper can simplify the speed control of locomotion. The new CPG model consisting of three layers can simulate the biological phenomenon and provide with a useful tool for researching the dynamical performance of locomotion.

Important Links:

Go Back