基于IC卡数据的公交乘客上下站点预测研究

Boarding and alighting stop inference based on transit IC card data

当前一票制的公交IC卡收费系统无法获取乘客的上车站点和下车站点,为公交线路客流量分析造成了一定困难。本文通过充分研究IC卡数据以及GPS数据,构建了以IC卡数据为基础的乘客上下车站点匹配模型,挖掘数据中包含的公交运行特征以及乘客出行特征。模型同时考虑封闭式公交出行链以及非封闭式公交出行链,并采用二分算法以及多概率融合的方法在乘客乘车特征基础上推测乘客的上下车站点,可有效提升公交的高效化运行管理。以承德市7号线路为例进行验证,采用线性回归方程对上下车站点匹配结果进行分析,相关系数高达0.977,校验回归方程系数值为0.9785。结果表明该模型可实现对公交车乘客上下车站点的有效匹配,具有较好的可靠性。

At present IC card charging system cannot obtain the passengers′boarding and alighting stop points which has caused certain difficulties for the analysis of the passenger flow of bus lines.By fully studying the IC card data and GPS data a passenger boarding and alighting station matching model based on IC card data is constructed to excavate the bus operation characteristics and passenger travel characteristics contained in the data.The model considers both the closed bus travel chain and the non-closed bus travel chain and uses the binary algorithm and multi-probability fusion method to speculate on the passenger′s boarding and alighting station based on the passenger′s riding characteristics which can effectively improve the efficient operation and management of the bus.Taking Chengde No.7 Line as an example for verification the matching results of the boarding and alighting stations are analyzed by using a linear regression equation and the correlation coefficient is as high as 0.977 and the regression equation coefficient value is 0.9785.The results show that the model can achieve effective matching of bus passengers boarding and alighting stops and have good reliability.