Adaptive Control for a Robotic Bipedal Device Using a Hybrid Discrete-Continuous Reinforcement Learning Strategy Academic Article in Scopus