成品油运输车轨迹频繁模式挖掘

Frequent Pattern Mining of Trajectory of Refined Oil Transport Vehicles

针对成品油运输车轨迹数据,提出一种结合轨迹聚类和地图匹配的频繁模式挖掘算法,该算法首先利用Fréchet距离进行相似性度量,再利用DBSCAN进行轨迹聚类,选取类簇中最长轨迹数据进行地图匹配,获取频繁货运路径。基于北京市部分运输车轨迹数据进行实验,获得实验结果进行分析,利用获得的频繁货运路径分析出成品油货运路径选择中安全便捷、低物流成本的特点。

A frequent pattern mining algorithm combining trajectory clustering and map matching is proposed for the trajectory data of refined oil transport vehicles.The algorithm first uses Fréchet distance for similarity measure,then uses DBSCAN for trajectory clustering,and selects the longest trajectory data in the class cluster for map matching to obtain frequent freight paths.Experiments were conducted based on the trajectory data of some transportation vehicles in Beijing,and the experimental results were obtained for analysis.The obtained frequent freight routes are used to analyze the characteristics of safety and convenience and low logistics cost in the freight route selection of refined oil products.