Unified walking control for a biped robot using neural networks

Abstract

A control scheme for handling variable speed gaits was implemented on an experimental biped. The control scheme used adaptation of pre-planned motion sequences in combination with closed loop reactive control. CMAC neural networks were responsible for the adaptive control of motion sequences. The biped was able to walk with variable speed gaits, and to change gait speeds on the fly. The slower gait speeds required statically balanced walking, while the faster speeds required dynamically balanced walking. It was not necessary to distinguish between the two balance modes within the controller. Following training, the biped was able to walk on flat, nonslippery surfaces at forward velocities in the range of 21 cm/min to 72 cm/min, with an average stride length of 6.5 cm.