车辆轨迹异常行为研究

Research on Abnormal Trajectory Behavior of Vehicles

摘要: 本文建立异常点检测模型，去除数据冗余后对异常数据进行数据清洗，即对轨迹数据漂移的部分选择性剔除。然后运用基于隐马尔科夫模型的地图匹配算法，结合实时电子地图，将轨迹偏移的一段匹配到与之最近的路网上，完成对运输车运输路线图的匹配，较好的纠正了车辆轨迹中的异常路段。

Abstract: In this paper, an outlier detection model is established. After removing data redundancy, the outlier data is cleaned up, that is partial selective removal of trajectory data drift. Then a map matching algorithm based on Hidden Markov Model and real-time electronic map are used to match a segment of the trajectory offset to the nearest road network, which completes the matching of the transportation road map of the transport vehicle and corrects the abnormal section of the vehicle trajectory better.