一种车辆GPS轨迹数据的改进地图匹配算法  

An Improved Map-matching Algorithm for GPS Trajectory Data of Vechiles

作　　者：文江辉[1,2] 霍佳琪 杨玲 WEN Jianghui;HUO Jiaqi;YANG Ling(School of Science,Wuhan University of Technology,Wuhan 430070,China;Intelligent Transport Systems Center,Wuhan University of Technology,Wuhan 430063,China)

机构地区：[1]武汉理工大学理学院,武汉430070 [2]武汉理工大学智能交通系统研究中心,武汉430063

出　　处：《武汉理工大学学报（交通科学与工程版）》2024年第1期1-6,12,共7页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国家自然科学基金(52272354);湖北省自然科学基金面上项目(2022CFB392)。

摘　　要：文中提出一种车辆GPS轨迹数据的改进地图匹配算法.以车辆GPS轨迹数据进行隐马尔可夫模型动态地图匹配后的缓冲区内候选路径为初始群体,以遗传算法中的改进的适应度函数为基准进行半交叉迭代,选择车辆轨迹更优的匹配路径;设计了改进地图匹配方法的算法流程;利用微软亚洲团队的GeoLift轨迹数据集验证了改进方法的有效性.An improved map matching algorithm for vehicle GPS trajectory data was proposed.Taking the candidate paths in the buffer after dynamic map matching of hidden Markov model with vehicle GPS trajectory data as the initial population,semi-crossover iteration was carried out based on the improved fitness function in genetic algorithm,and the better matching path of vehicle trajectory was selected,and the algorithm flow of improved map matching method was designed.The effectiveness of the improved method was verified by GeoLift trajectory data set of Microsoft Asia team.

关 键 词：GPS轨迹数据 隐马尔可夫地图匹配算法 改进地图匹配算法 改进遗传算法 

分 类 号：U495[交通运输工程—交通运输规划与管理]