基于批处理先验知识树的自主代客泊车路径规划  

Autonomous valet parking path planning based on batch informed tree

作　　者：张家旭 刘晔 郭崇[1] 赵健[1] ZHANG Jiaxu;LIU Ye;GUO Chong;ZHAO Jian(State Key Laboratory of Automotive Simulation and Control,Jilin University,Changchun 130022,China;Intelligent Network R&D Institute,China FAW Group Co.Ltd..Changchun 130011,China)

机构地区：[1]吉林大学汽车仿真与控制国家重点实验室,吉林长春130022 [2]中国第一汽车集团有限公司智能网联研发院,吉林长春130011

出　　处：《华中科技大学学报（自然科学版）》2021年第9期77-82,共6页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(51775235)。

摘　　要：针对室内结构化场景的自主代客泊车路径规划问题,基于批处理先验知识树(BIT*)算法和滚动优化思想提出一种自主代客泊车路径规划方法.首先,基于BIT*算法规划全局自主代客泊车路径,并采用数组实现影响BIT*算法搜索效率的树节点优先队列和树边优先队列,提高全局自主代客泊车路径规划效率.然后,基于滚动优化思想和圆弧-直线组合方法规划满足汽车运动学约束和避障约束的局部自主代客泊车路径,基于滚动优化思想规划的局部自主代客泊车路径可以将汽车引导到泊车位附近,而基于圆弧-直线组合方法规划的局部自主代客泊车路径可以使汽车安全驶入泊车位.最后,仿真验证所提出的自主代客泊车路径规划方法的可行性和有效性,结果表明:相对于基于传统快速扩展随机数(RRT)算法的自主代客泊车路径规划方法,提出的方法搜索效率更高,并且规划的自主代客泊车路径满足汽车运动学约束和避障约束.Aiming at the autonomous valet parking path planning problem for the indoor structured scenario,a novel autonomous valet parking path planning method was proposed based on batch informed tree algorithm and rolling optimization idea.Firstly,the global autonomous valet parking path was planned based on batch informed tree algorithm,and the tree node priority queue and the tree edge priority queue,which affected the search efficiency of batch informed tree algorithm,were implemented by using the array to improve the efficiency of the global autonomous valet parking path planning.Then,the local autonomous valet parking path meeting the vehicle kinematic constraint and the obstacle avoidance constraint was planned based on the rolling optimization idea and the arc-line combination method.The local autonomous valet parking path based on the rolling optimization idea could guide the vehicle to the neighborhood of the parking space,and the local autonomous valet parking path based on the arc-line combination method could make the vehicle drive into the parking space safely.Finally,the feasibility and effectiveness of the proposed method were verified by simulation.The results show that compared with the traditional rapidly exploring random tree-based autonomous valet parking path planning method,the proposed method has higher search efficiency,and the planned path meets both the vehicle kinematic constraint and the obstacle avoidance constraint.

关 键 词：自主代客泊车 路径规划 批处理先验知识树算法 滚动优化思想 圆弧直线组合方法 

分 类 号：U461.1[机械工程—车辆工程]