寒地城市地铁客流特性及网络失效研究

为分析寒地城市地铁客流与网络特性，挖掘客流与寒地气候的相关性，本文利用《寒区城市多模式公交协同运营技术与示范工程》2013—2020 年共 8 年 11 万余条地铁客流数据进行研究。提出效用阻抗Space L-Space P模型建立地铁抽象网络，将时间维度细分为周、月和年这3类， 研究地铁客流特性与扰动因素，并建立寒地城市地铁网络失效模型，分析扰动后地铁站点与线路的客流分布。进一步选取哈尔滨市与南京市地铁数据，运用转移熵因果关系分析地铁网络客流分布与气候的相关性，得到寒地气候对地铁客流的影响。研究表明：所采集的地铁客流数据能够充分展现客流的状态与变化趋势，能够满足客流数据分析的精度与质量要求。近8年，哈尔滨市地铁客流呈现明显的增长趋势，地铁客流分布呈现2月为客流低峰，3月客流逐步回升，3~12月客 流较平稳且具有轻微波幅；换乘客流中夏季呈现8月大于7月大于6月的趋势，冬季则2月换乘客流最低，大多在周一与周五达到周换乘客流量极值。由地铁网络失效模型识别出关键站点，得到 哈尔滨市冬季客流略高于其余3季，冬季地铁网络较脆弱，结合实际数据的模拟分析表明，温度与地铁客流具有一定相关性，且哈尔滨市的相关性大于南京市。

Subway Passenger Flow Characteristics and Network Failure Study for Cold Cities

This paper analyzes the passenger flow and network characteristics of subway in cold cities and explores the correlation between passenger flow and cold climate. More than 110000 subway passenger flow data from 2013 to 2020 from the "Multi-modal Public Transport Cooperative Operation Technology and Demonstration Project for Cold Cities" are used for the study. The paper proposes an effect impedance Space L-Space P model to establish the subway abstraction network. The time dimension is divided into week, month and years for the subway passenger flow characteristics and perturbation factors analysis. A failure model of the subway network is established for the cold cities, and the passenger flow distribution of subway stations and lines after the disturbance is analyzed. The subway data of Harbin and Nanjing cities are used to analyze the correlation between subway network passenger flow distribution and climate through the transfer entropy causality. The effect of cold climate on subway passenger flow is then obtained. The result shows that the collected subway passenger flow data can fully show the status and change trend of passenger flow, and meet the requirements of accuracy and quality of passenger flow data analysis. In the past eight years, the subway passenger flow in Harbin has shown an obvious growing trend. The distribution of subway passenger flow has a low peak in February, a gradual increase in March, and a stable and slightly fluctuating flow from March to December. In the summer, the transferring passenger flow decreases from August to June, and in winter, the lowest transferring passenger flow is observed in February. The weekly transferring passenger flow mostly reaches the peaks on Monday and Friday. The key stations were identified by the subway network failure model. It was found that the winter passenger flow in Harbin is slightly higher than other three seasons. And the subway network is more fragilein winter. The simulation analysis combined with the actual data shows that temperature and subway passenger flow have a certain correlation, and the correlation is more significant in Harbin than in Nanjing.