基于出租车数据的城市功能区分析——以西安市主城区为例  

Analysis of Urban Functional Areas Based on Cab Travel Data--An Example of the Main City of Xi’an

作　　者：贺泽宇 年雁云[1] 陈思文 边瑞 HE Zeyu;NIAN Yanyun;CHEN Siwen;BIAN Rui(College of Earth and Environmental Sciences,Lanzhou University,Lanzhou 730000,China)

机构地区：[1]兰州大学资源环境学院,甘肃兰州730000

出　　处：《城市勘测》2022年第4期17-22,共6页Urban Geotechnical Investigation & Surveying

基　　金：中国科学院十三五信息化项目(XXH13505-06)。

摘　　要：城市中不同区域的不同功能影响着人们的出行特征。城市交通配备的全球定位系统可以记录城市人口的时空特征,为探索居民的出行模式和城市功能区布局提供了可靠的数据源。本次研究以出租车数据、Point of interest(POI)数据为基础,将西安市主城区划分为3518个1 km×1 km的网格,计算2019年5月中旬一周的西安市出租车上、下车平均时间序列,并度量时间序列之间的相似性,利用k-medoids聚类方法得到不同类别并通过分析其间的差异来探索不同功能区中居民出行的时间特征,以此识别城市的功能布局。结果表明:对出租车行车数据的挖掘可以快速、准确识别城市功能区,了解居民出行模式,有利于城市管理和城市突发事件中的人员流动限制,对城市交通的统筹也有一定的参考意义;POI数据的加入提高了识别精度,弥补了出租车数据缺乏的城市功能细节。The different functions of different areas in the city influence the travel characteristics of people. Urban transportation equipped with GPS can record the spatial and temporal characteristics of urban population, providing a reliable data source for exploring the travel patterns of residents and the layout of urban functional areas. Based on cab data and POI data, this paper divides the main urban area of Xi’an into 3518 1 km×1 km grids, calculates the average time series of cab boarding and alighting in Xi’an for a week in mid-May 2019,and measures the similarity between the time series, uses k-medoids clustering method to get different categories and explores the temporal characteristics of residents’ trips in different functional areas by analyzing the differences between them, so as to identify the functional layout of the city. The results show that the mining of cab travel data can quickly and accurately identify urban functional areas and understand residents’ travel patterns, which is beneficial to urban management and the restriction of people’s movement in urban emergencies, and also has some reference significance for the coordination of urban traffic.In addition, the addition of POI data improves the recognition accuracy and makes up for the lack of cab data in the city function details.

关 键 词：出租车数据 出行模式 时间序列聚类 功能区识别 时空特征 

分 类 号：P208[天文地球—地图制图学与地理信息工程]