Design and Implementation of Fuzzy Policy Gradient Gait Learning Method for Walking Pattern Generation of Humanoid Robots

The design and implementation of Fuzzy Policy Gradient Learning (FPGL) method for humanoid robot is proposed in this paper. This paper not only introduces the phases of the humanoid robot walking, but also improves and parameterizes the gait pattern of the robot. FPGL is an integrated machine learning method based on Policy Gradient Reinforcement Learning (PGRL) and fuzzy logic concept in order to improve the efficiency and speed of gait learning computation. The result of the experiment shows that FPGL method can train the gait pattern from 9.26 mm/s walking speed to 162.27 mm/s within an hour. The training data of experiments also shows that this method could improve the efficiency of basic PGRL method up to 13%. The effect of arm movement to reduce the tilt of the trunk is also proved by the experimental results. All the results successfully demonstrate the feasibility and the flexibility of the proposed method.