基于轨迹数据的定制公交动态站点设置优化方法  

Optimization Method of Customized Bus Dynamic Station Setting Based on Trajectory Data

作　　者：巩方圆 刘林忠[1] GONG Fangyuan;LIU Linzhong(School of Traffic and Transportation,Lanzhou Jiaotong University,Lanzhou 730070,China)

机构地区：[1]兰州交通大学交通运输学院,兰州730070

出　　处：《综合运输》2023年第10期96-102,共7页China Transportation Review

基　　金：国家自然科学基金项目(71971019)。

摘　　要：合理预设定制公交的动态站点,可以满足乘客的就近乘降以及提高车辆载客率,同时也为定制公交线路规划、出行信息服务提供了依据和支撑。本文提出一种定制公交动态站点设置的优化方法,包括出租车轨迹数据提取出行热点区域和定制公交动态站点识别两部分。在出租车轨迹数据预处理的基础上,以乘客步行距离最小和动态站点数量最少为核心思想,建立一种基于改进DBSCAN算法的动态站点设置方法。最后通过选取某市某一时间段的出租车轨迹数据进行验证,结果表明,改进DBSCAN算法提高了适用性和准确率,降低了对领域知识的依赖性;提出的定制公交动态站点设置优化方法为定制公交站点的设置提供了依据。Reasonable preset dynamic stations of customized bus can meet the passengers'nearby ride and drop and improve the vehicle carrying rate,and also provide the basis and support for customized bus route planning and travel information services.This paper proposes an optimization method for the setting of customized bus dynamic stations,including taxi trajectory data extraction of travel hotspots and customized bus dynamic station identification.Based on the preprocessing of taxi trajectory data,a dynamic station setting method based on improved DBSCAN algorithm is established with the core idea of minimizing the walking distance of passengers and the number of dynamic stations.Finally,the taxi trajectory data of a certain time period in a city are selected for verification.The results show that the improved DBSCAN algorithm improves the applicability and accuracy and reduces the dependence on domain knowledge.The proposed optimization method of customized bus dynamic station setting provides a basis for the setting of customized bus stations.

关 键 词：定制公交 出租车轨迹数据 动态站点 DBSCAN算法 K-MEANS算法 

分 类 号：U491[交通运输工程—交通运输规划与管理]