摘要
为满足城市居民“最后一公里”出行需求,创新提出城市微循环公交潜在需求挖掘与线路设计算法。使用加权DBSCAN对共享单车、出租车、网约车等非公交出行OD进行聚类,挖掘出微循环线路潜在需求集中区域;通过分析历史出行数据揭示公交与共享单车竞争关系,构建以行程距离和公交相对共享单车行程时间优势为影响因素的公交客流估计模型,进而构建考虑客流收入和运营成本的线路收益估算模型,使用深度优先算法获取区内所有合理线路,从中选取收益最大线路作为推荐微循环线路。在广州市主城区范围内挖掘得到若干微循环公交潜在需求集聚区域,并选取其中典型区域计算微循环推荐线路,分析结果表明,推荐线路具备较强合理性。
To serve the"last mile"travel demand,a community shuttle potential demand extraction method and a recommended route design method are proposed.First,a weighted DBSCAN is applied to cluster the non-transit OD pairs and acquire regions where the need for a community bus route is substantial.Then,the competitive relationship between buses and shared bikes is revealed by analyzing historical trip data.A community bus OD demand estimation model with travel distance and relative travel time as factors is built.A route profit estimation model considering fee income and operation cost is proposed.A customized depth-first search is used to search for all reasonable routes.The route with the highest profit is chosen for recommendation.The proposed methods are applied to the main area of Guangzhou,and several potential high-demand areas are discovered.One typical area is selected to perform the shuttle route design method,the results show that the recommended route is highly reasonable.
作者
罗建平
黄子敬
张燕忠
LUO Jianping;HUANG Zijing;ZHANG Yanzhong(Guangdong Province Urban Intelligent Transportation Internet of Things Engineering Technology Research Center,Guangzhou 510000,China;Guangzhou Jiaoxintou Technology Co.,Ltd,Guangzhou 510000,China)
出处
《综合运输》
2023年第9期102-109,共8页
China Transportation Review
基金
广东省交通运输行业重点科技项目(2021-QD-012)
广州市科技计划项目(202206010056)。
关键词
城市交通
城市公交
微循环公交
数据挖掘
公交线路设计
Urban transportation
Urban transit
Community shuttle
Data mining
Transit route design
