| 摘要: |
| 铰接式车辆（AAV）由一辆动力牵引车与几辆无动力挂车组成，车辆在狭窄空间中运动时，挂车数量增多将导致避碰约束规模增加，并降低铰接式车辆运动规划任务的实时性。为此，提出了一种引导搜索与渐次优化相融合的铰接式车辆运动规划方法。首先，推导了双轴铰接式车辆通用运动学模型与规划任务约束条件；其次，考虑环境特性与铰接式车辆行为特征，对经典混合A*算法进行改进，引导搜索方向向目标节点扩展，从而生成符合铰接式车辆运动特性的初始路径；最后，以初始路径为热启动参考解，采用内点法（IPM）进行优化，将障碍物尺寸收缩后渐次扩展至原尺寸，并构建一系列避碰约束子问题以加速求解过程。仿真实验结果表明，引导搜索算法在拖曳不同级数挂车时均能够将求解效率提升40%~50%，且生成路径平滑性更优；渐次优化算法可通过减少引导搜索算法生成初始解中存在的小幅后向运动与航向角突变，以提升路径质量。 |
| 关键词: 运动规划 铰接式车辆 狭窄空间 混合A*算法 内点法 |
| DOI： |
|
| 基金项目:国家重点研发计划（2022YFC2603600）；国家自然科学基金（62233002） |
|
| Motion planning for two-axle articulated autonomous vehicle in narrow space with integration of guided search and progressive optimization |
| LI Daiwei,GAO Liang,JIA Bobo,XIE Shanshan,YANG Yi,FU Mengyin |
| (School of Automation, Beijing Institute of Technology, Beijing 100081, China; National Key Lab of Autonomous Intelligent Unmanned Systems, Beijing Institute of Technology, Beijing 100081, China;School of Automation, Beijing Institute of Technology, Beijing 100081, China; National Key Lab of Autonomous Intelligent Unmanned Systems, Beijing Institute of Technology, Beijing 100081, China; School of Automation, Nanjing University of Science and Technology, Nanjing 210094, China) |
| Abstract: |
| Articulated autonomous vehicle (AAV) consists of a powered tractor and several unpowered trailers. When the AAV maneuvers in narrow space, the increment of trailer amount leads to an increase in the scale of the collision avoidance constraints, and reduces the real-time capability of the motion planning tasks. To address this, an AAV motion planning method that combines guided search and progressive optimization is proposed. Firstly, the general kinematic model and planning task constraints for a two-axle AAV are derived. Secondly, the classical hybrid A* algorithm is improved by considering the characteristics of the environment and the AAV behaviour, and the initial path that matches the characteristics of the AAV motion is generated by guiding the search direction towards the target node. Finally, the interior point method (IPM) is used for the optimization, with the initial path used as the reference solution for the warm start. After shrinking the obstacle size, it is gradually expanded to the original size, and a series of collision avoidance constraint subproblems are constructed to speed up the solving process. Simulation experiment results show that the guided search algorithm can improve the planning efficiency by 40%~50% when towing different amount of trailers, and the smoothness of the generated path is better. The progressive optimization algorithm can improve the path quality by mitigating the small amount of backward maneuver and sudden changes of heading angle in the initial solution generated by the guided search algorithm. |
| Key words: Motion planning Articulated autonomous vehicle (AAV) Narrow environment Hybrid A* algorithm Interior point method (IPM) |
