The invention provides a parking task allocation and trajectory planning system based on multi-agent reinforcement learning, which comprises a parking space allocation subsystem and a trajectory planning subsystem, and is characterized in that vehicles are accessed to the parking space allocation subsystem after entering a parking lot environment, and corresponding parking spaces can be allocated to multiple vehicles at the same time, so that the time for finding the parking spaces after the vehicles enter is avoided, and the parking space allocation efficiency is improved. The parking efficiency and safety are improved; when the parking space distribution subsystem carries out parking space distribution, the demand of task distribution is embodied in reward value design, the shortest path reward and surrounding parking space occupation influence reward are designed at the same time, and the vehicle parking track length and the influence condition of surrounding environment congestion after parking are considered; and for the trajectory planning subsystem, when the optimal trajectory is obtained, the distance between the current position of the vehicle and the nearest obstacle is taken as one of consideration factors, the collision scene of the trajectory is further optimized, and on the premise of conforming to the dynamic constraint, not only is the intelligence of decision-making ensured, but also the driving safety is ensured.
本发明提供一种基于多智能体强化学习的泊车任务分配与轨迹规划系统,包括车位分配子系统和轨迹规划子系统,车辆进去停车场环境后接入车位分配子系统,支持多车同时分配对应的停车位,从而避免车辆在进入后寻找车位的时间,提升泊车效率和安全性;在车位分配子系统进行车位分配时,将任务分配的需求体现在奖励值设计中,同时设计最短路径奖励以及周围车位占据影响奖励,并考虑车辆泊车轨迹长度和泊车后周围环境阻塞的影响情况;对于轨迹规划子系统,在获取最优轨迹时,将车辆当前位置与距离最近的障碍物之间的距离作为考虑因素之一,进一步优化轨迹的碰撞场景,在符合动力学约束的前提下,既保证了决策的智能性,又确保了驾驶的安全性。
Parking task allocation and trajectory planning system based on multi-agent reinforcement learning
基于多智能体强化学习的泊车任务分配与轨迹规划系统
2023-08-22
Patent
Electronic Resource
Chinese
Trajectory Planning for Automated Parking Systems Using Deep Reinforcement Learning
Springer Verlag | 2020
|European Patent Office | 2024
|Parking path planning method based on reinforcement learning
European Patent Office | 2020
|Multi-vehicle cooperative trajectory planning parking optimization method
European Patent Office | 2024
|