出租车轨迹的同位模式挖掘算法  

Co-location Pattern Mining Algorithms for Taxi Trajectories Data

作　　者：徐诗奕 吴静[1,2,3] 傅优杰 XU Shiyi;WU Jing;FU Youjie(School of Surveying and Geoinformation Engineering,East China University of Technology,330013,Nanchang,PRC;Key Laboratory of Mine Environmental Monitoring and Improving around Poyang Lake of Ministry of Natural Resources,East China University of Technology,330013,Nanchang,PRC;Jiangxi Province Engineering Research Center of Surveying,Mapping and Geographic Information,330025,Nanchang,PRC;Eighth Geological Team of Jiangxi Geological Bureau,334000,Shangrao,Jiangxi,PRC)

机构地区：[1]东华理工大学测绘与空间信息工程学院,南昌330013 [2]东华理工大学自然资源部环鄱阳湖区域矿山环境监测与治理重点实验室,南昌330013 [3]江西省测绘地理信息工程技术研究中心,南昌330025 [4]江西省地质局第八地质大队,江西上饶334000

出　　处：《江西科学》2025年第2期385-390,共6页Jiangxi Science

基　　金：国家自然科学基金项目(41601416)。

摘　　要：出租车轨迹数据具有一定复杂性,现有的空间同位模式挖掘算法并不适用于轨迹数据集的挖掘。提出一种考虑时空轨迹时空关系和几何结构的同位模式挖掘算法。首先,将轨迹数据结构存储为数据流结构(Origin-Destination,OD);其次,将轨迹的时间、空间邻近性及方向相似性与时空邻域的计算相结合;最后,对南昌市2021年9月6日至12日的出租车轨迹数据集计算轨迹同位模式的频繁程度,从而挖掘不同时间段出租车司机在南昌站、南昌西站和昌北国际机场之间出行模式。结果表明,该方法能够准确挖掘出任何时间段的出租车司机在南昌站、南昌西站和昌北国际机场之间5种全局时空轨迹同位模式,并能有效分析3种全时间段的出行模式。Taxi trajectory data exhibits a certain degree of complexity,and the existing spatial co-location pattern mining algorithms are not suitable for trajectory datasets.This paper proposes a co-location pattern mining algorithm that considers the spatiotemporal relationships and geometric structures of trajectories.First,the trajectory data structure is stored in an OD(Origin-Destination)data stream structure.Second,the temporal and spatial proximity and directional similarity of trajectories are combined with the computation of spatiotemporal neighborhoods.Finally,the co-location pattern frequency of trajectories is calculated using a taxi trajectory dataset from Nanchang city,covering the period from September 6 to 12,2021.The algorithm is used to identify travel patterns between Nanchang Station,Nanchang West Station and Changbei International Airport at different times of the day.The results show that the method can accurately identify five global spatiotemporal co-location patterns among taxi drivers traveling between these three locations for any period,and effectively analyze three travel patterns across all time periods.

关 键 词：出租车轨迹数据 OD数据流 时空同位模式 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]