基于行程时间阈的城轨乘客乘车方案多维匹配推定模型

Deduction Method of Passengers’ Train Choices in Rail Transit Network Based on Travel Time Thresholds

摘要:

基于城市轨道交通网络化运营过程中的自动售检票系统（AFC）票卡和列车自动监控系统（ATS）行车数据以及乘客走行时间参数数据，提取乘客乘车方案行程时间阈，分析乘客实际选择的影响因素，提出基于行程时间阈的乘客乘车方案多维匹配推定模型。该模型实现了逐位票卡乘客推定至最可能的乘客方案上，并具有较强的合理性和实用性。最后通过北京城市轨道交通网络实际案例验证了方法的可行性和有效性。

Abstract:

Based on the automatic fare collection （AFC） data， automatic train supervision （ATS） data， and the passenger walking time data， travel time thresholds of train choices as well as the influence factors are analyzed in detail. Then a hierarchy deduction model of passengers’ train choices is developed. The model is able to accurately deduce passengers’ train choices one by one， Case studies on the Beijing metro network verifies the good performance of the model.