Abstract: A new way of collecting passenger flow information comes into being since the intelligent cards are used in public transport charge systems. The data recorded by these intelligent cards can help transport service providers obtain a large amount of statistic data they need. Confirming the boarding station of bus IC card holders is studied through the existing IC card data. At first, the hierarchical clustering method is adopted to classify the source data according to the minimum time interval between two passengers using IC card to get on the bus in adjacent stations. Then, the prediction value of the bus arrival time at each stop can be obtained from the bus schedule information. On the basis, the boarding station of bus IC card holders can be confirmed by the time matching between the results from IC card data and the bus arrival time at each stop. Finally, the feasibility and practicality of the method is proved by actual examples.

Key words: information of bus IC , hierarchical clustering method , forecast of bus arrival time ;

摘要： 公交IC卡是一种新的客流信息采集手段,它的广泛使用为公交管理者进行客流分析提供了新的数据源。充 分利用公交IC卡刷卡时间信息,根据相邻两站点间的最短刷卡间隔,通过系统聚类法对IC卡刷卡数据按照时间顺 序进行聚类分析。在此基础上,将聚类分析的结果与公交运营资料所推算出的车辆到站时间进行时间匹配,确定 持卡乘客的上车站点。最后,通过实例验证了该方法的可行性与实用性。

关键词: 公交IC卡信息, 系统聚类法, BUS到站时刻预测