Reinforcement learning method-based stable gait synthesis for biped robot | IEEE Conference Publication | IEEE Xplore