基于网约车订单数据的需求响应公交站点规划  

Demand responsive transit station planning based on online ride-hailing order data

作　　者：郭文举 吴芳[1] GUO Wenju;WU Fang(School of 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：《交通科技与经济》2025年第1期15-22,共8页Technology & Economy in Areas of Communications

基　　金：国家自然科学基金项目(42364003)。

摘　　要：为提高需求响应公交系统效率,精准满足乘客需求,缓解高峰期公交车乘客舒适度低和打车难问题,利用能够充分反应乘客出行需求的网约车订单数据,提出一种需求响应公交站点的规划方法。首先,通过对预处理的订单数据进行需求分析,挖掘交通需求出行特征和出行习惯。其次,利用改进的DBSCAN算法对数据进行聚类分析,筛选出热点区域,挖掘交通需求的空间特征。在此基础上,利用改进的K-means算法划分高需求区域,得到每个热点区域中心点,将相邻中心点间距较近的点进行整合并规划站点。最后,通过实例验证其可靠性和科学性。结果表明,通过该方法规划的站点更具实用性和灵活性,可更好地满足乘客出行需求,提高需求响应公交系统的吸引力和服务效率,缓解交通拥堵。In order to improve the efficiency of demand responsive transit system,accurately meet the needs of passengers,and alleviate the problems of low comfort of bus passengers and difficulty in taking a taxi during peak hours,a demand responsive transit station planning method is proposed by using the online ride-hailing order data that can fully reflect the travel needs of passengers.Firstly,through the demand analysis of the pre-processed order data,the travel characteristics and travel habits of traffic demand are mined.Secondly,the improved DBSCAN algorithm is used to perform cluster analysis on the data,screen out the hot spots,and mine the spatial characteristics of traffic demand.On this basis,the improved K-means algorithm is used to divide the high demand area,and the center point of each hot spot area is obtained.The points with close distance between adjacent center points are integrated and the site is planned.Finally,its reliability and scientificity are verified by examples.The results show that the stations planned by this method are more practical and flexible,which can better meet the travel needs of passengers,improve the attractiveness and service efficiency of the demand responsive transit system,and alleviate traffic congestion.

关 键 词：交通工程 站点规划 DBSCAN算法 K-MEANS算法 需求响应公交 网约车订单数据 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]