基于数据融合的公共自行车接驳公交行为辨识

城市智能交通卡和公交自动车辆定位系统的大规模应用为研究公共自行车换乘公交的出行规律提供了契机。本文结合多源数据，研究公共自行车接驳公交的换乘时空阈值，根据自行车站点与公交站点间距离构建匹配站点对；并基于时间阈值提取匹配站点对的刷卡记录构建换乘识别模型的训练集和测试集，应用关联规则挖掘算法实现同一用户公共自行车IC卡卡号与公交IC卡号的匹配；根据设定的验证规则，92%的匹配规则通过验证。之后基于换乘识别结果分析公共自行车接驳常规公交的换乘客流的时空特性。结果表明：公共自行车接驳常规公交呈现出明显的早晚高峰时段，且接驳出行主要分布在城市中心区域。

Behavior identification of public bicycle access regular bus transit based on data fusion

The large-scale application of urban intelligent transportation card and bus GPS vehicle orientation system provides an opportunity to study the travel rules of public bicycle transfer to public transport. Based on the fusion of multi-source data, the transfer time-space threshold of public bicycle access public transit were determined and the matching station pair according to the distance between the bicycle station and the public transportation station were constructed; then the training set and the test set of the recognition model were constructed, the association rule mining algorithm was applied to realize the matching between the IC card number of public bicycle and the IC card number of public transportation for the same user; 92% of the matching rules are validated, proving the effectiveness of the matching rules. Then, based on the transfer identification results, the spatial and temporal characteristics of the transfer passenger flow between public bicycles and regular buses are analyzed. The results show that the public bicycles access bus transit trips present obvious morning and evening peak hours, and the trips are mainly distributed in the urban center.