【摘要】 在地下车库排布车位，往往受到车库轮廓、障碍物和连通性等条件约束，本文设计并实现了一种基于探索策略和区域划分的车位排布方案。探索策略借鉴了强化学习的思想，通过设置奖励机制使智能体在地下车库环境中进行主路的铺设；区域划分算法可以在保证不堵塞车道情况下得到尽可能多的车位数量。本文算法能够在短时间内获得车位排布结果，帮助设计师减轻工作量，提高项目收益。更多还原

【Abstract】 In underground garages, parking spaces are often constrained by garage contours, obstacles, connectivity and other conditions. In this paper, a parking arrangement scheme based on exploration strategy and regional division is designed and implemented. The exploration strategy draws on the idea of reinforcement learning, and makes the agent lay the main road in the underground garage environment by setting up a reward mechanism. The regional division algorithm can get as many parking spaces as possible without blocking the lanes. The proposed algorithm can obtain the results of parking arrangement in a short time, which can help designers reduce workload and improve project income.更多还原