基于自动售检票数据的城市轨道交通通勤客流辨识
Commuters identification for urban rail transit using automatic fare collection data

摘要 城市轨道交通具有明显的通勤客流主体特征,把握通勤客流出行规律对运营管理具有重要意义.本文以海量自动售检票数据潜在包含的时空关系为基础,从时间、空间、个体属性、出行规律四方面构建基于规则的出行目的辨识算法,重点针对通勤(上下班及上下学)客流进行辨识.以2014年北京市轨道交通售检票数据进行实证分析,结果显示:该方法能有效辨识上班、上学、下班回家、放学回家、其他回家和其他6类客流,其中上班占比26.77%,上学占0.44%,回家占44.49%(包含下班回家、放学回家及其他回家),其他占28.30%.结合2014年北京市公共交通出行调查结果,验证了辨识结果的准确性.该研究扩展了售检票数据应用范围,为精细化客流特征研究提供了一种低成本、高效的分析方法.
Abstract：The passenger flow of Urban Rail Transit (URT) mainly composes of commuters, and it is of great significance to understand the law of travel characteristics of commuters for the operation and management. Based on the potential spatial-temporal relationship within Automatic Fare Collection (AFC) records, a rule-based identification method is proposed to infer trip purpose for URT travelers, which is constructed from the view of time, space, personal properties and regular travel behaviors. An empirical transit network from Beijing in China is applied to verify the efficiency of the proposed method, results show that six types of trips (home-work, work-home, home-school, school-home, others-home and other) can be efficiently identified, where the work trips cover 26.77 %, the school trips cover 0.44 %, home trips cover 44.49 % and other trips cover 28.30 %. Compared to the public transportation survey results in Beijing of 2014, the identification results are verified to be reasonable and acceptable. This study improves the value of AFC data, and provides a new method for deeply analyzing travel demand.