369 / 2024-03-22 14:50:25

Reinforcement Learning-Driven Heuristic Path Planning Method for Automated Special Vehicles in Unstructured Environment

Unstructured rugged terrain,Automated special vehicles (ASVs),Global path planning,A* algorithm,Reinforcement learning (RL)

4、智能化装备技术 > 3、特殊环境机器人技术

Aiming at improving the adaptability of global path planning method for the automated special vehicles (ASVs) in a variety of unstructured environments, a reinforcement learning (RL)-driven heuristic path planning method is proposed. The introduction of traditional heuristic algorithm avoids inefficiency of RL in the early learning phase, and it provides a preliminary planning path to be adjusted by RL. Meanwhile, a reward function is designed based on vehicle dynamics to generate a smooth, stable, and efficient path. The simulation environment is established based on real terrain data. The algorithm performance is tested through setting different starting and ending points, and conducting seven cases. And the distribution of different obstacles on the path planning influence of the ASVs is discussed. Results verify that the proposed method can get a collision free and efficient path with excellent adaptability to complex terrains.