苏月同,徐天捷,蒲一超,许项东

• 1:同济大学
• 2:道路与交通工程教育部重点实验室
• 3:上海申通地铁集团有限公司技术中心

摘要(Abstract)：

识别城市轨道交通站点高峰时段，对合理分配站内管理资源、制定乘客限流和错峰出行方案，从而缓解线路站点的高峰拥挤现象等具有重要作用。在现有多数城市的实践和研究中，主要依据人工经验确定全网或单条线路固定长度的高峰时段，但随着城市轨道网络规模和客流的增长，该方法难以体现不同站点和线路高峰时段的差异性，为车站开展精细化运营管理带来了挑战。针对城市轨道交通网络中的每个站点，本文基于以5 min为单元的进出站连续客流数据，提出了一种基于有序样本聚类的站点级差异化高峰时段识别方法。根据识别结果，进一步定义高峰时段时间窗最大客流、峰左（右）客流比和高峰时段长度三个指标，将网络中的站点高峰分为无高峰、微弱高峰、明显高峰三类。最后，以上海轨道交通18条运营线路5个工作日的客流数据为例，验证了方法的有效性。分析结果表明：(1)所提出方法可同时辨识出高峰时段的开始时刻和结束时刻，无须预先确定高峰时段长度，并且针对高峰时段的特点，使用定制化聚类参数，能够识别全网各站点差异化高峰时段；(2)同一条线路中站点距市中心越远，其进站早高峰时段开始越早，验证了辨识差异化高峰时段的必要性。

关键词(KeyWords)： 城市轨道交通;进出站客流;高峰时段;差异化识别;有序样本聚类

Abstract:

Keywords:

基金项目(Foundation): 上海市青年科技启明星计划项目（20QA1409800）

作者(Author): 苏月同,徐天捷,蒲一超,许项东

DOI: 10.19961/j.cnki.1672-4747.2022.10.007

参考文献(References)：

• [1]中国政府网.国家综合立体交通网规划纲要[EB/OL].(2021-02-24)[2022-9-28]. http://www.gov.cn/gongbao/content/2021/content_5593440.htm.
• [2]张雨洁,李宗平.大客流条件下地铁旅客运输组织[J].交通运输工程与信息学报, 2017, 15(2):58-63,70.ZHANG Yu-jie, LI Zong-ping. Subway passenger transport organization under the condition of large passenger flow[J]. Journal of Transportation Engineering and Information, 2017, 15(2):58-63,70.
• [3]宋美艳,王亮,朱沁萱.基于逐时客流量的地铁站空调系统动态负荷研究[J].西南科技大学学报, 2022, 37(2):58-63.SONG Mei-yan, WANG Liang, ZHU Qin-xuan. Study on dynamic load of air-conditioning systems in subway stations based on hour-by-hour passenger flow[J]. Journal of Southwest University of Science and Technology, 2022,37(2):58-63.
• [4] HALVORSEN A, KOUTSOPOULOS H N, LAU S, et al.Reducing subway crowding:analysis of an off-peak discount experiment in Hong Kong[J]. Transportation Research Record:Journal of the Transportation Research Board, 2016, 2544(1):38-46.
• [5]余丽洁.城市轨道交通车站高峰客流偏移下的客流预测理论与方法研究[D].西安：长安大学, 2020.YU Li-jie. Theory and method of passenger flow prediction based on urban rail transit station peak deviation[D].Xi’an:Chang’an University, 2020.
• [6]李星毅.基于相似性的交通流分析方法[D].北京：北京交通大学, 2010.LI Xing-yi. Methods of traffic flow analysis based on similarity[D]. Beijing:Beijing Jiaotong University, 2010.
• [7]王静,刘剑锋,孙福亮.北京市轨道交通线网客流分布及成长规律[J].城市交通, 2012, 10(2):26-32.WANG Jing, LIU Jian-feng, SUN Fu-liang. Passenger demand distribution and increasing trend over Beijing transit network[J]. Urban Transportation of China, 2012, 10(2):26-32.
• [8]成艳,叶霞飞,王治,等.城市轨道交通高峰时段站间起讫点矩阵预测模型[J].同济大学学报（自然科学版）,2018, 46(3):346-353.CHENG Yan, YE Xia-feng, WANG Ye, et al. Forecasting model of peak-period station-to-station origin-destination matrix in urban rail transit systems[J]. Journal of Tongji university(Natural Science), 2018, 46(3):346-353.
• [9]张晚笛,陈峰,王子甲,等.基于多时间粒度的地铁出行规律相似性度量[J].铁道学报, 2018, 40(4):9-17.ZHANG Wan-di, CHEN Feng, WANG Zi-jia, et al. Similarity measurement of metro travel rules based on multitime granularities[J]. Journal of the China Railway Society, 2018, 40(4):9-17.
• [10]金昱.上海市轨道交通车站客流高峰特征研究[J].城市交通, 2019, 17(4):50-57.JIN Yu. Characteristics of peak hour passenger flow at rail transit stations in Shanghai[J]. Urban Transportation of China, 2019, 17(4):50-57.
• [11]顾丽萍,叶霞飞.轨道车站客流高峰时段与土地利用关系研究[M].品质交通与协同共治——2019年中国城市交通规划年会论文集,北京:中国建筑工业出版社, 2019:1087-1100.
• [12]石庄彬.城轨交通客流高峰事件持续时间预测[D].南京:东南大学, 2015.SHI Zhuang-bin. Estimation of congestion duration on urban rail transit[D]. Nanjing:Southeast University,2015.
• [13]王文宪,肖蒙,成琳娜,等.基于近邻传播聚类的地铁运营时段划分[J].运筹与管理, 2018, 27(12):187-192.WANG Wen-xian, XIAO meng, CHENG Lin-na, et al.Classification of subway operation intervals based on affinity propagation cluster[J]. Operations Research and Management Science, 2018, 27(12):187-192.
• [14]陈东洋,陈德旺,江世雄,等.基于时段客流特征聚类的地铁运营时段划分[J].计算机系统应用, 2021, 30(3):256-261.CHEN Dong-yang, CHEN De-wang, JIANG Shi-xiong,et al. Division of metro operation periods based on feature clustering of passenger flow[J]. Computer Systems&Applications, 2021, 30(3):256-261.
• [15]曾小旭,汪林,罗贤迪,等.有序样本聚类方法在城市轨道交通运营时段划分中的应用[J].都市快轨交通,2017, 30(2):108-112.ZENG Xiao-xu, WANG Lin, LUO Xian-di, et al. Application of oridinal clustering in the division of operation periods for urban rail transit[J]. Urban Rapid Rail Transit, 2017, 30(2):108-112.
• [16]陈宽民,余丽洁,马超群.西安市城市轨道交通车站高峰时段偏差研究[J].城市交通, 2018, 16(5):51-58, 7.CHEN Kuan-min, YU Li-jie, MA Chao-qun. Differentiated peak hours at urban rail transit stations in Xi’an[J].Urban Transportation of China, 2018, 16(5):51-58,7.
• [17] BULUT A, YANIK S. Partitioning the time of day for bus schedule optimization[C]//KAHRAMAN C, CEBI S, CEVIK ONAR S, et al. INFUS 2019:Intelligent and Fuzzy Techniques in Big Data Analytics and Decision Making. Istanbul:Springer International Publishing.2020:298-304.
• [18] WANG X, ZHANG Y, LIU H, et al. An improved robust principal component analysis model for anomalies detection of subway passenger flow[J]. Journal of Advanced Transportation, 2018, 2018:e7191549.
• [19] GU J, JIANG Z, FAN W, et al. Real-time passenger flow anomaly detection considering typical time series clustered characteristics at metro stations[J]. Journal of Transportation Engineering Part A:Systems, 2020, 146(4):04020015.
• [20]方开泰.有序样品的一些聚类方法[J].应用数学学报,1982,5(1):94-101.FANG Kai-tai. Some clustering methods for the order sample[J]. Acta Mathematicae Applicatae Sinica, 1982(1):94-101.
• [21]王昌,姜仙童,吴艳平.基于聚类分析的公交线路资源配置时段划分方法——以烟台市为例[J].交通运输研究, 2021, 7(5):35-42.WANG Chang, JIANG Xian-tong, WU Yan-ping. Time division method of bus route resource allocation based on cluster analysis:a case study of Yantai[J]. Transport Research, 2021, 7(5):35-42.
• [22]张雪妍,贺锋,王玉刚,等.城市公交系统运营时段准确划分研究与仿真[J].计算机仿真, 2019, 36(11):126-129.ZHANG Xue-yan, HE Feng, WANG Yu-gang, et al. Research and simulation on operation period division of urban bus system[J]. Computer Simulation, 2019, 36(11):126-129.
• [23]刘乔.对传统聚类与模糊聚类选取标准的分析研究[D].北京:中国地质大学(北京), 2012.LIU Qiao. Analysis and research on selection standards of classic and fuzzy cluster methods[D]. Beijing:China University of Geosciences(Beijing), 2012.
• [24]马士宾,虞秋富,张静,等.基于有序聚类法的大中修路段划分技术研究[J].河北工业大学学报, 2014, 43(5):106-109.MA Shi-bin, YU Qiu-fu, ZHANG Jing, et al. Technical research on pavement maintenance division based on ordered sample clustering[J]. Journal of Hebei University of Technology, 2014, 43(5):106-109.
• [25]马跃欣.基于有序聚类的Wi-Fi指纹室内定位算法研究[D].太原:太原理工大学, 2020.MA Yue-xin. Research on Wi-Fi fingerprint indoor positioning algorithm based on ordinal cluster[D]. Taiyuan:Taiyuan University of Technology, 2020.
• [26]孙伯才.改进有序聚类方法在水文分析中的应用[J].水土保持应用技术, 2021(5):19-21.
• [27]杜树樱.高速公路交通检测器布设优化方法研究[D].南京:东南大学, 2019.DU Shu-ying. Optimal methods of traffic sensors placement on highways[D]. Nanjing:Southeast University,2019.
• [28]中国城市规划设计研究院. 2022年度中国主要城市通勤监测报告[R].北京:中国城市规划设计研究院,2022.