Learning dynamic balance of a biped walking robot

This paper discusses the application of CMAC (cerebellar model arithmetic computer) neural networks to the problem of biped walking with dynamic balance. The project goal is to develop biped control strategies based on a hierarchy of simple gait oscillators, PID controllers and neural network learning, but requiring no detailed dynamic models. The focus of this report is on real-time control studies using a ten axis biped robot with joint position, foot force and body acceleration sensors. While efficient walking has not yet been achieved, the experimental biped has learned the closed chain kinematics necessary to shift body weight from side-to-side while maintaining good foot contact and has learned the dynamic balance required in order to lift a foot off the floor for a desired length of time, during which the foot can be moved to a new location relative to the body. Using these skills, the biped is able to link short steps without falling.<<ETX>>