北京市出租车运量分布的时空格局及生成机制

doi:10.11821/dlyj020200652

摘要/Abstract

摘要：

出租车是城市交通的重要组成部分,对其精细化的管控需要理解运量分布的时空特征和生成机制。通过北京市出租车大数据,采用系统聚类法将其15分钟时间片段归并为空间分布相似的时段,刻画出运量分布的时空格局,进而采用地理加权回归模型分析了生成机制。本研究揭示的出租车运量分布变化的时间点并不完全对应传统的整时点;各时段运量均呈现空间集聚的特征,但集聚的位置、面积明显不同,工作日不同时段运量集聚程度的差异较周末更大;不同时段出租车运量的影响因子有所差异,其中商业设施、房价、地铁和公交站密度、道路密度等因子通常较为显著。研究结果对时空上更精细的出租车运量预测、出租车分区分时段的政策管制和规划管理具有启示意义。

关键词: 出租车, 轨迹数据, 系统聚类法, 地理加权回归, 时空特征

Abstract:

Taxi is one of the important components of the urban transportation system, either insufficient or excessive supply can have an adverse impact on city development. Therefore, the industry needs to be reasonably controlled. Targeted measures for taxis requires an in-depth understanding of spatiotemporal characteristics and local associated factors of taxi ridership. With the help of taxi GPS data, our study slices time at a finer scale than the existing research to 15 minutes. Based on the spatial distribution of taxi ridership within each time slice, the 96 time slices are merged by the hierarchical clustering method. Then the geographically weighted regression (GWR) model is used to analyze the local associated factors for the generation mechanism of taxi traffic ridership during major periods of weekdays and weekends. The study area is within the Fifth Ring Road in Beijing. The conclusions are as follows: (1) Taxi ridership varies greatly between weekdays and weekends, and the spatial distribution changes rapidly during morning peak hours. (2) The timespans clustered are roughly similar to those in the existing research, but this paper reveals 8:15, 7:30, and other non-integral period divisions, which are not showed in research based on traditional peak hour or 1-hour time scale. (3) The local associated factors vary significantly in different areas and at different timespans, but the influence of commercial facilities, house price, subway and bus station density, and road density is usually prominent. The density of commercial facilities makes a great contribution in most areas; contribution from house price is generally negative in high-price areas; the density of subway stations and bus stations usually have positive contribution to taxi ridership; the density of primary roads has a greater impact on pick-ups, while that of secondary roads influences drop-offs more. These results suggest that control policies by time and by region are needed when we make refined policies for the taxi industry in the future.

Key words: taxi, trajectory data, hierarchical clustering, GWR, spatiotemporal characteristics

施念邡, 杨星斗, 戴特奇. 北京市出租车运量分布的时空格局及生成机制[J]. 地理研究, 2021, 40(6): 1667-1683.

SHI Nianfang, YANG Xingdou, DAI Teqi. The spatiotemporal pattern of taxi ridership and its generation mechanism in Beijing[J]. GEOGRAPHICAL RESEARCH, 2021, 40(6): 1667-1683.

图/表 20

图1

图2

表1

表2

图3

表3

图4

表4

图5

表5

图6

表6

表7

图7

图8

图9

图10

表8

图11

图12

参考文献 34

[1]	Liu Q, Ding C, Chen P. A panel analysis of the effect of the urban environment on the spatiotemporal pattern of taxi demand. Travel Behaviour and Society, 2020, 18: 29-36. DOI: 10.1016/j.tbs.2019.09.003. doi: 10.1016/j.tbs.2019.09.003
[2]	李岩. 基于热点区域的出租车出行需求预测模型研究. 西安: 长安大学硕士学位论文, 2019: 1.
	[Li Yan. Research on taxi travel demand forecasting model based on hot areas. Xi'an: Master Dissertation of Chang'an University, 2019: 1.]
[3]	郭锐欣, 毛亮. 特大城市出租车行业管制效应分析: 以北京市为例. 世界经济, 2007, (2): 75-83.
	[Guo Ruixin, Mao Liang. An analysis of the regulatory effect of taxi industry in megacities: A case study of Beijing. The Journal of World Economy, 2007, (2): 75-83] DOI: 10.3969/j.issn.1002-9621.2007.02.009. doi: 10.3969/j.issn.1002-9621.2007.02.009
[4]	杨振山, 龙瀛, DOUAY Nicolas. 大数据对人文-经济地理学研究的促进与局限. 地理科学进展, 2015, 34(4): 410-417. doi: 10.11820/dlkxjz.2015.04.002
	[Yang Zhenshan, Long Ying, Douay Nicolas. Opportunities and limitations of big data applications to human and economic geography: The state of the art. Progress in Geography, 2015, 34(4): 410-417.] DOI: 10.11820/dlkxjz.2015.04.002. doi: 10.11820/dlkxjz.2015.04.002
[5]	张博, 邓浩坤, 安宁, 等. 大数据背景下国内人文地理研究进展与方法审视. 人文地理, 2019, 34(3): 14-21, 90.
	[Zhang Bo, Deng Haokun, An Ning, et al. The re-examination of human geography research and the research methods in the context of big data. Human Geography, 2019, 34(3): 14-21, 90.] DOI: 10.13959/j.issn.1003-2398.2019.03.003. doi: 10.13959/j.issn.1003-2398.2019.03.003
[6]	Liu Y, Wang F, Xiao Y, et al. Urban land uses and traffic “source-sink areas”: Evidence from GPS-enabled taxi data in Shanghai. Landscape and Urban Planning, 2012, 106(1): 73-87. DOI: 10.1016/j.landurbplan.2012.02.012. doi: 10.1016/j.landurbplan.2012.02.012
[7]	程静, 刘家骏, 高勇. 基于时间序列聚类方法分析北京出租车出行量的时空特征. 地球信息科学学报, 2016, 18(9): 1227-1239. doi: 10.3724/SP.J.1047.2016.01227
	[Cheng Jing, Liu Jiajun, Gao Yong. Analyzing the spatio-temporal characteristics of Beijing's OD trip volume based on time series clustering method. Journal of Geo-Information Science, 2016, 18(9): 1227-1239.] DOI: 10.3724/SP.J.1047.2016.01227. doi: 10.3724/SP.J.1047.2016.01227
[8]	刘菊, 许珺, 蔡玲, 等. 基于出租车用户出行的功能区识别. 地球信息科学学报, 2018, 20(11): 1550-1561. doi: 10.12085/dqxxkx.2018.180164
	[Liu Ju, Xu Jun, Cai Ling, et al. Identifying functional regions based on the spatio-temporal pattern of taxi trajectories. Journal of Geo-Information Science, 2018, 20(11): 1550-1561.] DOI: 10.12085/dqxxkx.2018.180164. doi: 10.12085/dqxxkx.2018.180164
[9]	庄浩铭, 刘小平. 基于出租车轨迹数据的城市空间结构变化研究: 以深圳市为例. 热带地理, 2020, 40(2): 217-228.
	[Zhuang Haoming, Liu Xiaoping. Identifying changes in urban spatial structure using taxi trajectory data: A case study in Shenzhen. Tropical Geography, 2020, 40(2): 217-228.] DOI: 10.13284/j.cnki.rddl.003224. doi: 10.13284/j.cnki.rddl.003224
[10]	塔娜, 曾屿恬, 朱秋宇, 等. 基于大数据的上海中心城区建成环境与城市活力关系分析. 地理科学, 2020, 40(1): 60-68. doi: 10.13249/j.cnki.sgs.2020.01.008
	[Ta Na, Zeng Yutian, Zhu Qiuyu, et al. Relationship between built environment and urban vitality in Shanghai downtown area based on big data. Scientia Geographica Sinica, 2020, 40(1): 60-68.] DOI: 10.13249/j.cnki.sgs.2020.01.008. doi: 10.13249/j.cnki.sgs.2020.01.008
[11]	周勍, 秦昆, 陈一祥, 等. 基于数据场的出租车轨迹热点区域探测方法. 地理与地理信息科学, 2016, 32(6): 51-56, 127.
	[Zhou Qing, Qin Kun, Chen Yixiang, et al. Hotspots detection from taxi trajectory data based on data field clustering. Geography and Geo-Information Science, 2016, 32(6): 51-56, 127.] DOI: 10.3969/j.issn.1672-0504.2016.06.009. doi: 10.3969/j.issn.1672-0504.2016.06.009
[12]	曲昭伟, 王鑫, 宋现敏, 等. 基于出租车GPS大数据的城市热点出行路段识别方法. 交通运输系统工程与信息, 2019, 19(2): 238-246.
	[Qu Zhaowei, Wang Xin, Song Xianmin, et al. Urban hotspot travel section identification method based on taxi GPS large data. Journal of Transportation Systems Engineering and Information Technology, 2019, 19(2): 238-246.] DOI: 10.16097/j.cnki.1009-6744.2019.02.034. doi: 10.16097/j.cnki.1009-6744.2019.02.034
[13]	孙飞, 张霞, 唐炉亮, 等. 基于GPS轨迹大数据的优质客源时空分布研究. 地球信息科学学报, 2015, 17(3): 329-335. doi: 10.3724/SP.J.1047.2015.00329
	[Sun Fei, Zhang Xia, Tang Luliang, et al. Temporal and spatial distribution of high efficiency passengers based on GPS trajectory big data. Journal of Geo-Information Science, 2015, 17(3): 329-335.] DOI: 10.3724/SP.J.1047.2015.00329. doi: 10.3724/SP.J.1047.2015.00329
[14]	付鑫, 孙茂棚, 孙皓. 基于GPS数据的出租车通勤识别及时空特征分析. 中国公路学报, 2017, 30(7): 134-143.
	[Fu Xin, Sun Maopeng, Sun Hao. Taxi commute recognition and temporal-spatial characteristics analysis based on GPS data. China Journal of Highway and Transport, 2017, 30(7): 134-143.] DOI: 10.19721/j.cnki.1001-7372.2017.07.017. doi: 10.19721/j.cnki.1001-7372.2017.07.017
[15]	齐兰兰, 周素红, 闫小培. 广州市医疗设施可达性模型中端点吸引的影响因素检验. 地理科学, 2014, 34(5): 580-586. doi: 10.13249/j.cnki.sgs.2014.05.580
	[Qi Lanlan, Zhou Suhong, Yan Xiaopei. Endpoint attractive factors of medical facilities' accessibility: Based on GPS floating car data in Guangzhou. Scientia Geographica Sinica, 2014, 34(5): 580-586.] DOI: 10.13249/j.cnki.sgs.2014.05.580. doi: 10.13249/j.cnki.sgs.2014.05.580
[16]	陈卓然, 黄翀, 刘高焕, 等. 基于出租车GPS数据的居民就医时空特征分析. 地球信息科学学报, 2018, 20(8): 1111-1122. doi: 10.12082/dqxxkx.2018.180157
	[Chen Zhuoran, Huang Chong, Liu Gaohuan, et al. Analysis of spatial-temporal characteristics of resident travel for hospitals based on taxi GPS data. Journal of Geo-Information Science, 2018, 20(8): 1111-1122.] DOI: 10.12082/dqxxkx.2018.180157. doi: 10.12082/dqxxkx.2018.180157
[17]	王姣娥, 杜方叶, 靳海涛, 等. 基于交通出行链的就医活动识别理论框架与方法体系. 地球信息科学学报, 2020, 22(4): 805-815. doi: 10.12082/dqxxkx.2020.190566
	[Wang Jiaoe, Du Fangye, Jin Haitao, et al. Identifying hospital-seeking behavior based on trip chain data: Theoretical framework and methodological system. Journal of Geo-Information Science, 2020, 22(4): 805-815.] DOI: 10.12082/dqxxkx.2020.190566. doi: 10.12082/dqxxkx.2020.190566
[18]	杨文军, 孟凡冬, 李游, 等. 基于车辆轨迹OD数据分析乘客出行特征的研究. 测绘与空间地理信息, 2020, 43(S1): 144-147, 150.
	[Yang Wenjun, Meng Fandong, Li You, et al. Research on passenger travel characteristics based on OD data analysis of vehicle trajectory, Geomatics & Spatial Information Technology, 2020, 43(S1): 144-147, 150.] DOI: 1672-5867(2020)S0-0144-04. doi: 1672-5867(2020)S0-0144-04
[19]	刘安, 宋伟芳, 杨佩昆. 出行生成预测模型的研究. 同济大学学报: 自然科学版, 1998, (3): 290-293.
	[Liu An, Song Weifang, Yang Peikun. Study on trip generation forecasting. Journal of Tongji University: Natural Science Edition, 1998, (3): 290-293.] DOI: CNKI: SUN: TJDZ.0.1998-03-019. doi: CNKI: SUN: TJDZ.0.1998-03-019
[20]	石飞, 王炜, 陆建. 居民出行生成预测方法的归纳和创新. 城市交通, 2005, 3(1): 43-46.
	[Shi Fei, Wang Wei, Lu Jian. Improvement and conclusion about resident trip generation. Urban Transport of China, 2005, 3(1): 43-46.] DOI: 10.3969/j.issn.1672-5328.2005.01.011. doi: 10.3969/j.issn.1672-5328.2005.01.011
[21]	林雄斌. 面向拥堵治理的交通政策创新及其精细化管理趋势. 道路交通与安全, 2016, 16(4): 1-6.
	[Lin Xiongbin. Recent innovations in transport policies and delicacy management trend. Road Traffic & Safety, 2016, 16(4): 1-6.] DOI: 10.13986/j.cnki.jote.2016.04.001. doi: 10.13986/j.cnki.jote.2016.04.001
[22]	Tu W, Cao R, Yue Y, et al. Spatial variations in urban public ridership derived from GPS trajectories and smart card data. Journal of Transport Geography, 2018, 69(MAY): 45-57. DOI: 10.1016/j.jtrangeo.2018.04.013. doi: 10.1016/j.jtrangeo.2018.04.013
[23]	谢蔚翰, 周素红. 建成环境对出租车出行需求影响的时空间分异模式. 现代城市研究, 2018, 33(12): 22-29.
	[Xie Weihan, Zhou Suhong. The spatial-temporal-nonstationary effect of built environment on taxi demand. Modern Urban Research, 2018, 33(12): 22-29.] DOI: 10.3969/j.issn.1009-6000.2018.12.004. doi: 10.3969/j.issn.1009-6000.2018.12.004
[24]	张自荷, 王振, 吴瑞. 基于空间滞后模型的出租车需求影响因素分析. 武汉理工大学学报: 交通科学与工程版, 2019, 43(2): 351-356.
	[Zhang Zihe, Wang Zhen, Wu Rui. Analysis of influential factors of taxi demand based on spatial lag model. Journal of Wuhan University of Technology: Transportation Science & Engineering, 2019, 43(2): 351-356.] DOI: 10.3963/j.issn.2095-3844.2019.02.034. doi: 10.3963/j.issn.2095-3844.2019.02.034
[25]	郝新华, 周素红, 柳林. 广州市高峰小时居住用地交通产生与分布: 基于出租车GPS起讫关联的分析. 人文地理, 2016, 31(6): 109-114, 151.
	[Hao Xinhua, Zhou Suhong, Liu Lin. Traffic generation and distribution of residential land of Guangzhou in commuting peak hour: based on the analysis of O-D associated data of taxi GPS. Human Geography, 2016, 31(6): 109-114, 15.] DOI: 10.13959/j.issn.1003-2398.2016.06.014. doi: 10.13959/j.issn.1003-2398.2016.06.014
[26]	刘萌, 邬群勇. 基于出租车OD数据的居民活动强度时空特征研究. 福州大学学报: 自然科学版, 2018, 46(2): 204-209.
	[Liu Meng, Wu Qunyong. Spatial and temporal patterns of residents activity intensity based on taxi OD data. Journal of Fuzhou University: Natural Science Edition, 2018, 46(2): 204-209.] DOI: 10.7631/issn.1000-2243.17016. doi: 10.7631/issn.1000-2243.17016
[27]	周博, 马林兵, 胡继华, 等. 基于轨迹数据场的热点区域提取及空间交互分析: 以深圳市为例. 热带地理, 2019, 39(1): 117-124.
	[Zhou Bo, Ma Linbing, Hu Jihua, et al. Extraction of urban hotspots and analysis of spatial interaction based on trajectory data field: A case study of Shenzhen city. Tropical Geography, 2019, 39(1): 117-124.] DOI: 10.13284/j.cnki.rddl.003097. doi: 10.13284/j.cnki.rddl.003097
[28]	Yoo E H. How short is long enough? Modeling temporal aspects of human mobility behavior using mobile phone data. Annals of the American Association of Geographers, 2019, 109(5): 1415-1432. DOI: 10.1080/24694452.2019.1586516. doi: 10.1080/24694452.2019.1586516
[29]	Li B, Cai Z, Jiang L, et al. Exploring urban taxi ridership and local associated factors using GPS data and geographically weighted regression. Cities, 2019, 87: 68-86. DOI: 10.1016/j.cities.2018.12.033. doi: 10.1016/j.cities.2018.12.033
[30]	Qian X, Ukkusuri S V. Spatial variation of the urban taxi ridership using GPS data. Applied Geography, 2015, 59: 31-42. DOI: 10.1016/j.apgeog.2015.02.011. doi: 10.1016/j.apgeog.2015.02.011
[31]	Zhang X, Xu Y, Tu W, et al. Do different datasets tell the same story about urban mobility: A comparative study of public transit and taxi usage. Journal of Transport Geography, 2018, 70: 78-90. DOI: 10.1016/j.jtrangeo.2018.05.002. doi: 10.1016/j.jtrangeo.2018.05.002
[32]	白同舟, 蔡乐, 朱家正, 等. 轨道交通与城市协同发展的空间差异性分析: 以北京市为例. 交通运输系统工程与信息 , 2020, 20(3): 14-19.
	[Bai Tongzhou, Cai Le, Zhu Jiazheng, et al. Spatial differentiation of rail transit and urban development coordination: A case study of Beijing. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(3): 14-19.] DOI: 10.16097/j.cnki.1009-6744.2020.03.003. doi: 10.16097/j.cnki.1009-6744.2020.03.003
[33]	薛冰, 肖骁, 李京忠, 等. 基于POI大数据的老工业区房价影响因素空间分异与实证. 人文地理, 2019, 34(4): 106-114.
	[Xue Bing, Xiao Xiao, Li Jingzhong, et al. POI-based analysis on the affecting factors of property prices' spatial distribution in the traditional industrial area. Human Geography, 2019, 34(4): 106-114.] DOI: 10.13959/j.issn.1003-2398.2019.04.013. doi: 10.13959/j.issn.1003-2398.2019.04.013
[34]	戴特奇, 金凤君, 张华, 等. 双交通模式的阿朗索模型对中国城市的模拟研究. 地理学报, 2010, 65(10): 1266-1274. doi: 10.11821/xb201010011
	[Dai Teqi, Jin Fengjun, Zhang Hua, et al. Urban equilibrium model with a bimodal transportation system in China. Acta Geographica Sinica, 2010, 65(10): 1266-1274.] DOI: 10.11821/xb201010011. doi: 10.11821/xb201010011

类别	自变量	单位
土地利用	住宅POI密度	个/km²
	商业设施POI密度	个/km²
	公司企业POI密度	个/km²
	公共服务POI密度	个/km²
	旅游景区POI密度	个/km²
	教育设施POI密度	个/km²
	交通设施POI密度	个/km²
	POI混合熵	-
	房价	元/m²
人口特征	人口密度	人/km²
	中年人口比例	-
	本科及以上学历比例	-
交通设施	地铁站密度	个/km²
	公交站密度	个/km²
	一级道路密度	km/km²
	二级道路密度	km/km²

工作日上车点	工作日下车点	周末上车点	周末下车点
23:45—6:30	0:00—5:45	0:00—7:00	23:30—7:30
6:30—8:15	5:45—7:30	7:00—12:00	7:30—10:30
8:15—9:45	7:30—16:00	12:00—0:00	10:30—19:30
9:45—23:45	16:00—0:00	-	19:30—23:30

工作日时段	Moran's I	Z-score	置信水平
23:45—6:30	0.335	29.010	0.99
6:30—8:15	0.456	34.687	0.99
8:15—9:45	0.527	39.291	0.99
9:45—23:45	0.463	34.756	0.99

工作日时段	Moran's I	Z-score	置信水平
0:00—5:45	0.472	35.547	0.99
5:45—7:30	0.091	10.661	0.99
7:30—16:00	0.291	23.741	0.99
16:00—0:00	0.479	35.953	0.99

周末时段	Moran's I	Z-score	置信水平
0:00—7:00	0.444	34.502	0.99
7:00—12:00	0.457	35.226	0.99
12:00—0:00	0.410	31.248	0.99