Abstract: In the case of complex parking scenarios, the existing parallel parking algorithms have some problems such as small scope of application, long computing time and high failure rate. In order to solve these problems, an automatic parking path planning algorithm based on geometric curves and improved hybrid A* is designed. Firstly, the kinematic model of the vehicle is established according to Ackermann's steering theory. Secondly, the geometric curve method is designed to plan the path of the simple parallel parking scene. Then, an improved hybrid A* is designed to improve the success rate of parking planning in complex parallel scenes. Finally, simulation software is used to carry out simulation analysis and real vehicle verification. The results show that the automatic parking path planning algorithm based on geometric curves and improved hybrid A* can effectively expand the application scope of parking environment, enhance computing power, improve the success rate of parallel parking planning, and has a very wide application prospect in the parking field.

Key words: automatic parking; parallel parking; geometric curve method; improved hybrid A* ; path planning

摘要： 在泊车场景复杂的情况下，现有的平行泊车算法存在环境适用范围小、计算时间长、 失败率高等问题，为了解决这一问题，设计了一种基于几何曲线和改进混合 A *的自动泊车路 径规划算法。首先，根据阿克曼的转向理论建立了汽车的运动学模型；其次，设计了几何曲 线法来对简单的平行泊车场景进行路径规划；然后，设计了一种改进混合 A *来提高复杂平行 场景泊车的规划成功率；最后，利用仿真软件进行仿真分析并实车验证。结果表明，基于几 何曲线和改进混合 A *的自动泊车路径规划算法能够有效地扩大泊车环境适用范围，增强计算 能力，提高平行泊车规划成功率，并在泊车领域具有非常广泛的应用前景。

关键词: 自动泊车；平行泊车；几何曲线法；改进混合 A *；路径规划