基于辅助信息校验的高准确隐马尔可夫地图匹配技术研究  

Research on High-Accuracy Hidden Markov Map-Matching Technology with Auxiliary Information Verification

作　　者：张华[1,2] 战兴群 吕翔宇[2] 王文杰 ZHANG Hua;ZHAN Xingqun;LV Xiangyu;WANG Wenjie(School of Aeronautics and Astronautics,Shanghai Jiao Tong University,Shanghai 200240,China;Jiaoxin Beidou(Hainan)Technology Co.,Ltd.,Haikou 570311,China)

机构地区：[1]上海交通大学航空航天学院,上海200240 [2]交信北斗(海南)科技有限公司,海口570311

出　　处：《交通工程》2025年第4期12-16,26,共6页Journal of Transportation Engineering

基　　金：科学技术部重点研发计划项目“广域交通时空大数据综合服务平台与应用示范”(2022YFB3904405)。

摘　　要：隐马尔可夫模型(HMM)地图匹配算法在网约车里程收费、物流车辆行驶路径规划、车辆保险评估等方面广泛应用。针对传统HMM算法在复杂路网中识别准确率下降的不足,提出一种优化算法,通过融合速度、方向及点位连续性等数据进行辅助校验,可显著提升复杂路网中的匹配准确性。基于海南真实路网数据和车辆轨迹数据的实验验证表明,优化算法具有更高的鲁棒性,能精准还原车辆轨迹,准确计量车辆行驶里程,可支撑海南里程收费等应用。Hidden Markov Model(HMM)map matching algorithm is widely used in network car mileage charging,logistics vehicle route planning,vehicle insurance evaluation and so on.To address the issue that the traditional HMM algorithm has a decreased recognition accuracy in complex road networks,this paper proposes an optimized algorithm.By integrating data on speed,direction,and point continuity for auxiliary verification,the algorithm can significantly improve the matching accuracy in complex road networks.Experiments based on real road network data and vehicle trajectory data from Hainan demonstrate that the optimized algorithm has higher robustness,can accurately reconstruct vehicle trajectories,and precisely measure vehicle travel distance,effectively supporting applications such as mileage billing in Hainan.

关 键 词：隐马尔可夫模型 北斗轨迹还原 地图匹配 里程计费 

分 类 号：U491[交通运输工程—交通运输规划与管理]