基于多源地理大数据与机器学习的地铁乘客出行目的识别方法

doi:10.12082/dqxxkx.2020.200134

摘要/Abstract

摘要：

探索地铁乘客出行目的识别方法,有助于突破智能卡数据（Smart Card Data,SCD）在具体应用场景中的局限性,提升SCD在交通出行研究、交通发展规划等领域的应用价值。本文融合多源地理大数据,基于城市交通与土地利用时空间互动理论,以北京市居民地铁出行为例,在交通出行调查数据中提取5565个地铁出行样本及其对应的出行目的和出行特征相关变量。基于兴趣点（Point of Interest,POI）数据得到各样本起止站点的土地利用特征相关变量,形成包含每次地铁出行的出行目的、出行特征、土地利用特征的地铁出行数据集。使用基于随机森林（Random Forest,RF）算法对地铁出行数据集进行训练完成的分类器对SCD记录的每一次地铁出行进行分类,获得该次出行的出行目的及其不同目的地铁出行时空间分布规律。研究结果表明,本识别方法可有效预测地铁乘客的出行目的,其中,“上班”、“回家”2类出行目的的预测准确率均超过90%;纳入土地利用特征相关变量可显著提升RF分类器预测准确率,印证了城市交通与土地利用的时空间互动理论。鉴于当前SCD的可获取性逐渐提高,该项技术在居民地铁出行监测与预测、地铁线网布局和地铁周边土地利用规划等实践方面,具有很强的推广性,有助于更全面地认知大城市居民的地铁出行行为。

关键词: 地铁出行, 出行目的识别, 交通调查数据, 智能卡数据, 兴趣点数据, 随机森林, 土地利用, 时空间互动, 北京

Abstract:

Identifying metro trip purpose using Smart Card Data (SCD) is important to expand the application of SCD in transport research and transport planning. This paper integrates different types of big data and combines the theories on the interaction between transport and land use. By taking Beijing as a case, we firstly analyze the metro trip purposes of individual passengers using travel survey data from 5565 respondents. Secondly, we investigate the land use features of trip origin and destination using Point of Interest(POI) data . Thirdly, a metro trip dataset is developed which includes the information of trip purpose, trip duration, and spatial distribution of trip origin and destination. Fourthly, a Random Forest (RF) algorithm is used to establish a RF classifier using the metro trip dataset as training data. Finally, this trained classifier is used to classify each metro trip recorded by the SCD to identify the metro trip purpose and the spatial distribution of metro trips for different purposes. The results of analysis show that the random forest classifier trained in this study can effectively identify metro trip purposes from SCD. For trips with "go to work" and "go home" purposes, the accuracy of identification can reach over 90%. One reason for the high identification accuracy is that land use information is included in the RF classifier. Our results confirm the theory of spatial-temporal interactions between transport and land use. There is an increasing availability of multi-source geographic big data and traffic survey data of residents in large cities, which means that the method developed in this study would have a high value in metro trip predicting and monitoring, transport planning, and land use policy-making around the metro stations. Also, our results enhance our knowledge of metro travel behavior in megacities.

Key words: Metro trips, trip purpose, travel survey data, smart card data, point of interest data, Random Forest algorithm, land use, spatial-temporal interactions, Beijing

赵鹏军, 曹毓书. 基于多源地理大数据与机器学习的地铁乘客出行目的识别方法[J]. 地球信息科学学报, 2020, 22(9): 1753-1765.DOI:10.12082/dqxxkx.2020.200134

ZHAO Pengjun, CAO Yushu. Identifying Metro Trip Purpose using Multi-source Geographic Big Data and Machine Learning Approach[J]. Journal of Geo-information Science, 2020, 22(9): 1753-1765.DOI:10.12082/dqxxkx.2020.200134

图/表 13

表1

表2

图1

图2

图3

表3

图4

图5

图6

表4

图7

图8

表5

参考文献 36

[1]	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[J]. 地球信息科学学报, 2020,22(4):805-815.
	[ Wang J E, Du F Y, Jin H T, et al. Identifying hospital-seeking behavior based on trip chain data: Theoretical framework and methodological system[J]. Journal of Geo-information Science, 2020,22(4):805-815. ]
[2]	Ali Safwat K N, Magnanti T L. A combined trip generation, trip distribution, modal split and trip assignment model[J]. Transportation Science, 1988,22(1):14-30. doi: 10.1287/trsc.22.1.14
[3]	Handy S. Methodologies for exploring the link between urban form and travel behavior[J]. Transportation Research Part D: Transport and Environment, 1996,1(2):151-165.
[4]	Maria Kockelman K. Travel behavior as function of accessibility, land use mixing, and land use balance: Evidence from San Francisco Bay Area[J]. Transportation research record, 1997,1607(1):116-125.
[5]	赵鹏军, 万婕. 城市交通与土地利用一体化模型的理论基础与发展趋势[J]. 地理科学, 2020,40(1):12-21.
	[ Zhao P J, Wan J. The key technologies of integrated urban transport-land use model: Theory base and development trends[J]. Scientia Geographica Sinica, 2020,40(1):12-21. ]
[6]	陆化普, 孙智源, 屈闻聪. 大数据及其在城市智能交通系统中的应用综述[J]. 交通运输系统工程与信息, 2015,15(5):45-52.
	[ Lu H P, Sun Z Y, Qu W C. Big data and its applications in urban intelligent transportation system[J]. Journal of Transportation Systems Engineering and Information Technology, 2015,15(5):45-52. ]
[7]	龙瀛, 张宇, 崔承印. 利用公交刷卡数据分析北京职住关系和通勤出行[J]. 地理学报, 2012,67(10):1339-1352.
	[ Long Y, Zhang Y, Cui C Y. Identifying commuting pattern of Beijing using bus smart card data[J]. Acta Geographica Sinica, 2012,67(10):1339-1352. ]
[8]	韩昊英, 于翔, 龙瀛. 基于北京公交刷卡数据和兴趣点的功能区识别[J]. 城市规划, 2016,40(6):52-60.
	[ Han H Y, Yu X, Long Y. Identifying urban functional zones using bus smart card data and points of interest in Beijing[J]. City Planning Review, 2016,40(6):52-60. ]
[9]	Zhang Y, Martens K, Long Y. Revealing group travel behavior patterns with public transit smart card data[J]. Travel Behaviour and Society, 2018,10:42-52.
[10]	Liu K, Yin L, Ma Z, et al. Investigating physical encounters of individuals in urban metro systems with large-scale smart card data[J]. Physica A: Statistical Mechanics and its Applications, 2019: 123398.
[11]	Bagchi M, White P R. The potential of public transport smart card data[J]. Transport Policy, 2005,12(5):464-474.
[12]	Pelletier M P, Trépanier M, Morency C. Smart card data use in public transit: A literature review[J]. Transportation Research Part C: Emerging Technologies, 2011,19(4):557-568.
[13]	Morency C, Trepanier M, Agard B. Measuring transit use variability with smart-card data[J]. Transport Policy, 2007,14(3):193-203.
[14]	Ma X, Wu Y J, Wang Y, et al. Mining smart card data for transit riders' travel patterns[J]. Transportation Research Part C: Emerging Technologies, 2013,36:1-12.
[15]	Long Y, Thill J C. Combining smart card data and household travel survey to analyze jobs-housing relationships in Beijing[J]. Computers, Environment and Urban Systems, 2015,53:19-35.
[16]	Kusakabe T, Asakura Y. Behavioural data mining of transit smart card data: A data fusion approach[J]. Transportation Research Part C: Emerging Technologies, 2014,46:179-191.
[17]	Ma X, Liu C, Wen H, et al. Understanding commuting patterns using transit smart card data[J]. Journal of Transport Geography, 2017,58:135-145.
[18]	赵鹏军, 李南慧, 李圣晓. TOD建成环境特征对居民活动与出行影响——以北京为例 [J]. 城市发展研究, 2016(6):45-51.
	[ Zhao P J, Li N H, Li S X. The impacts of the built environment on residents' acitivities and travel behavior in TOD areas: A case study of Beijing[J]. Urban Development Studies, 2016(6):45-51. ]
[19]	文萍, 赵鹏军, 周素红. TOD对居民通勤模式的影响:以英国泰恩威尔都市区为例, 国际城市规划, 2019(8):1-15.
	[ Wen P, Zhao P J, Zhou S H. Effects of TOD on residents' commuting patterns: Empirical evidence from tyne and wear, the UK[J]. Urban Planning International, 2019(8):1-15. ]
[20]	Cervero R. Transit-oriented development in the United States: Experiences, challenges, and prospects[C]. TCRP Report, 2004,102.
[21]	Zhao P, Zhang Y. The effects of metro fare increase on transport equity: New evidence from Beijing[J]. Transport Policy, 2019,74:73-83.
[22]	Cervero R, Day J. Suburbanization and transit-oriented development in China[J]. Transport Policy, 2008,15(5):315-323.
[23]	Zhao P, Li S. Bicycle-metro integration in a growing city: The determinants of cycling as a transfer mode in metro station areas in Beijing[J]. Transportation Research Part A: Policy and Practice, 2017(99):46-60.
[24]	Zhao P, Li P. Travel satisfaction inequality and the role of the urban metro system[J]. Transport Policy, 2019,79:66-81.
[25]	Olaru D, Smith B, Taplin J H E. Residential location and transit-oriented development in a new rail corridor[J]. Transportation Research Part A: Policy and Practice, 2011,45(3):219-237.
[26]	Li S, Zhao P. Exploring car ownership and car use in neighborhoods near metro stations in Beijing: Does the neighborhood built environment matter?[J]. Transportation research part D: Transport and Environment, 2017,56:1-17.
[27]	Zhao P, Li S. Suburbanization, land use of TOD and lifestyle mobility in the suburbs[J]. Journal of Transport and Land Use, 2018,11(1):195-215.
[28]	Alonso W. Location and land use: Toward a general theory of land rent[J]. Economic Geography, 1964,42(3):11-26.
[29]	Hansen W G. How accessibility shapes land use[J]. Journal of the American Planning Association, 1959,25(2):73-76.
[30]	赵鹏军, 孔璐. TOD对北京市居民通勤影响及其机制研究[J]. 人文地理, 2017,32(5):125-131.
	[ The impact of TOD on residents' commuting activities: A case of Beijing. Human Geography, 2017,32(5):125-131. ]
[31]	Breiman L. Random forests[J]. Machine learning, 2001,45(1):5-32. doi: 10.1023/A:1010933404324
[32]	Liaw A, Wiener M. Classification and regression by random Forest[J]. R news, 2002,2(3):18-22.
[33]	曹泽涛, 方子东, 姚瑾, 等. 基于随机森林的黄土地貌分类研究[J]. 地球信息科学学报, 2020,22(3):452-463.
	[ Cao Z T, Fang Z D, Yao J, et al. Loess landform classification based on random forest[J]. Journal of Geo-information Science, 2020,22(3):452-463. ]
[34]	方匡南, 吴见彬, 朱建平, 等. 随机森林方法研究综述[J]. 统计与信息论坛, 2011,26(3):32-38.
	[ Fang K N, Wu J B, Zhu J P, et al. A review of technologies on random forests[J]. Statistics & Information Forum, 2011,26(3):32-38. ]
[35]	Han H, Guo X L, Yu H. Variable selection using Mean Decrease Accuracy and Mean Decrease Gini based on Random Forest[C].IEEE International Conference on Software Engineering & Service Science. IEEE, 2016: 219-224.
[36]	Townsend J T. Theoretical analysis of an alphabetic confusion matrix[J]. Perception & Psychophysics, 1971,9(1):40-50.

数据类型	数据描述	数据年份	数据来源
居民出行调查数据	居民一日出行链	2015年（对应2014年北京市居民出行情况）	北京市交通委员会（http://jtw.beijing.gov.cn/）
SCD智能卡数据	共计约1434万条地铁出行数据	2018年（7月1日至7月7日）	北京市交通委员会（http://jtw.beijing.gov.cn/）
百度POI数据	用于反映城市服务设施的空间分布情况	2015年	百度地图开放平台（http://lbsyun.baidu.com/）
地铁站点数据	北京市地铁站点空间分布情况	2014年、2018年	北京地铁（https://www.bjsubway.com/）
住房交易价格数据	单位面积成交价格	2015年	北京链家网（https://bj.lianjia.com/）

出行目的	样本数量/条	占比/%
回家	3270	58.76
其他	498	8.95
上班	1797	32.29
总计	5565	100.00

特征名称	特征描述
出行目的	被识别变量（上班、回家、其他）
出行特征	出发时刻、到达时刻、出行时长
土地利用特征	起止点周边高收入、低收入工作场所类型POI核密度值
	起止点周边居民点类型兴趣点与住房价格核密度值
	起止点周边公共服务与生活服务设施类型POI核密度值
	起止点到市中心欧式距离

	样本数量/条			样本占比/%
	分类结果为“回家”	分类结果为“其他”	分类结果为“上班”	预测准确样本占比
真实值为“回家”	782	38	14	93.76
真实值为“其他”	8	45	23	59.21
真实值为“上班”	0	42	439	91.27

	样本数量/条			样本占比/%
	分类结果为 “回家”	分类结果为 “其他”	分类结果为 “上班”	仅包含出行特征分类准确样本占比	初始RF分类器准确样本占比
真实值为“回家”	765	38	15	93.52	93.76
真实值为“其他”	22	31	33	36.05	59.21
真实值为“上班”	3	56	428	87.89	91.27