基于电子收费数据的车辆轨迹匹配

为充分利用ETC门架采集的数据对高速公路运营管理提供支撑，实现车辆轨迹地图匹配，本文结合图论理论，基于车辆ETC门架数据设计Floyd算法还原车辆行驶路径。首先，对路径匹配系统需求进行分析，介绍ETC联网收费系统的组成和功能、讨论ETC门架系统的组成、布局以及工作流程。其次，分析车辆行驶ETC门架数据，提取与研究相关的字段信息并进行数据预处理，从而将门架信息匹配到了实际路网地图上。第三，设计车辆路径匹配的算法，在路径信息不完整的情况下，根据Floyd算法得出的最短路径矩阵，使用matlab运行程序，得到拟合的车辆行驶全程路径。最后，开展多场景下的基于ETC数据的车辆轨迹匹配案例验证。分别按照车辆经过的门架信息数据完整、经过门架信息数据在高速公路互通处缺失、不在互通处缺失、数据大量缺失四种类型的场景下，使用算法得到车辆行驶的完整路径，完成了多场景下车辆轨迹的地图匹配，从而实现后续的车辆完整路径下的计费。

Research on Vehicle Trajectory Matching Based on ETC Data

In order to make full use of the data collected by ETC gantry to provide support for expressway operation and management and realize vehicle trajectory map matching, this paper combines graph theory and designs Floyd algorithm based on vehicle ETC gantry data to restore vehicle driving path. Firstly, the requirements of the path matching system were analyzed, the composition and function of the ETC network charging system were introduced, and the composition, layout and work flow of the ETC portal system were discussed. Secondly, in order to match the mast information to the actual road network map, the data of the ETC mast of the vehicle were analyzed, the field information related to the research were extracted and data were preprocessed. Thirdly, the vehicle path matching algorithm were designed. In the case of incomplete path information, the shortest path matrix was obtained according to the Floyd algorithm, and the fitted vehicle path was obtained by running the matlab software program. Finally, vehicle trajectory matching case verification based on ETC data in multiple scenarios is carried out. According to the four types of scenarios where the mast information data passed by the vehicle was complete, the mast information data was missing at the expressway interchange, not at the interchange, and a large amount of data was missing, the algorithm was used to obtain the complete path of the vehicle. The map of the vehicle trajectory in the scene was matched, so the subsequent charging under the complete path of the vehicle was realized.