A Reinforcement Learning Based Dynamic Walking Control

Abstract

A quasi-passive dynamic walking robot is built to study natural and energy-efficient biped walking. The robot is actuated by MACCEPA actuators. A reinforcement learning based control method is proposed to enhance the robustness and stability of the robot's walking. The proposed method first learns the desired gait for the robot's walking on a flat floor. Then a fuzzy advantage learning method is used to control it to walk on uneven floor. The effectiveness of the method is verified by simulation results.