基于特征站点的公交IC卡数据站点匹配方法研究

    Approach on Station ID and Trade Record Match Based on Characteristic Stations in Bus IC Data Mining

    • 摘要: 设计了一种基于特征站点的公交IC卡数据站点匹配方法,首先利用k-means聚类方法,通过计算交易时间间隔将同一站点产生的交易数据归类,再根据线路站点客流特征和乘客出行的换乘行为寻找线路特征站点(客流峰值站点与换乘站点),站点匹配时优先将特征站点编号与聚类数据进行匹配,从而提高匹配精度.在数据实验中,将该方法与基于站点间运行时间的匹配方法进行了对比,结果显示,基于特征站点的匹配方法平均准确率为85%.

       

      Abstract: A station ID match method based on characteristic stations is developed in this approach.According to the trade time interval calculation,the k-means clustering algorithm is used to gather the trade records which occur in the same stations.The characteristic stations including maximum passenger volume station and transfer station are confirmed by analyzing the history IC data.The characteristic stations have priority in station ID match to achieve the higher accuracy.For comparison,the method based on link travel time match is used in data experiment,and the results shows the accuracy of characteristic stations achieves to 85%.

       

    /

    返回文章
    返回