基于时空分析的地图匹配算法研究  

Map-matching algorithm research based on spatio-temporal analysis

作　　者：刘卫宁[1] 汪杰宇 郑林江[1,2,3] 

机构地区：[1]重庆大学计算机学院,重庆400044 [2]重庆城市综合交通枢纽开发投资有限公司,重庆401121 [3]中铁二院重庆勘察设计研究院有限责任公司,重庆400023

出　　处：《计算机应用研究》2016年第8期2266-2269,共4页Application Research of Computers

基　　金：国家自然科学基金资助项目(61203135);中国博士后科学基金资助项目(2014T70852);重庆市博士后科研资助项目(XM201305);重庆市应用开发计划重点项目(cstc2014yykf B30003);中央高校基金重大项目(106112014CDJZR188801)

摘　　要：针对目前城市浮动车数据量日益庞大,地图匹配算法实时性差、匹配率不高的缺点,提出了一种基于时空分析的地图匹配算法。算法在对城市路网建立网格索引的基础上,综合考虑了空间几何、路网拓扑信息及上下文因素对选取GPS投影点的影响,大大提高了匹配效率和匹配精度。实验结果表明,算法能够满足工程应用中浮动车地图匹配的实时性和准确性。Considering the large volumes of floating car data, the lower real-time and map matching rate of map matching algo- rithm in current urban transportation, this paper proposed a map-matching algorithm based on spatio-temporal analysis. On the basis of grid index building for urban road network, this algorithm considered the spatial geometric, topological structures info of the road network and the impact on the selection of GPS point＇ s projection, which greatly enhanced the matching efficiency and improved the matching accuracy. The experiment proves that this algorithm can meet the requirement of real-time performance and accuracy of floating car map matching in the practical application.

关 键 词：地图匹配 路网拓扑 网格索引 时空分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]