网格索引支持下的大规模浮动车实时地图匹配方法

发布时间：2018-04-26 11:09

本文选题：浮动车技术 + 地图匹配　；参考：《计算机辅助设计与图形学学报》2014年09期

【摘要】：针对大规模浮动车数据量庞大、地图匹配时对实时性要求高的特点,提出一种离线计算与在线匹配相结合的浮动车实时局部地图匹配方法.在离线计算部分,首先对对矢量道路网进行二级网格化分,然后对二级网格的每一个栅格,以其中心点坐标到各个候选路段或道路节点的距离以及候选路段的通行方向计算距离和方向的组合相似度,根据组合相似度最大值与次大值之差的大小分成4种情形,并将组合相似度明显占优的候选路段或者道路节点作为该栅格的属性值预先存储;在线匹配时,针对实时获取的轨迹点,采用文中提出的"双向前行-后行法"快速判定前后时刻欲匹配路段之间的路径可达度,并据此判断所属情形;当满足条件时,可立即获得当前轨迹点的匹配结果,即该轨迹点所在栅格的属性值.实验结果表明了该文方法处理海量浮动车数据的可行性和有效性:通过对深圳市大规模浮动车实测数据进行验证,该方法比局部地图匹配方法的单点匹配平均耗时减少了1倍;对整个深圳市区构建两级网格并存储属性值,其占有存储空间约为8 MB,可以将其一次性读入内存;整个深圳市道路网的离线计算时间约需花费32 h,但离线计算一经完成就可以长期服务于实时匹配.
[Abstract]:In view of the large amount of data of large floating vehicle and the high requirement of real time in map matching, a real time local map matching method for floating vehicle is proposed, which combines off-line calculation and online matching. In the off-line calculation part, the vector road network is first divided into two levels of mesh, and then each grid of the secondary grid is divided. Based on the distance from the center coordinate to each candidate section or road node and the combination similarity between the distance and direction of the candidate road, it is divided into four cases according to the magnitude of the difference between the maximum and the second largest of the combination similarity. The candidate section or road node with obvious similarity is stored as the attribute value of the grid in advance, and when online matching, the locus points are obtained in real time. In this paper, the "bidirectional forward-back method" is used to quickly determine the path reachability between the sections to be matched before and after the time, and to judge the case accordingly. When the conditions are satisfied, the matching results of the current locus points can be obtained immediately. The property value of the grid in which the trace point is located. The experimental results show that the method is feasible and effective in dealing with the massive floating vehicle data. Through the verification of the measured data of the large-scale floating vehicle in Shenzhen, the average time consuming of this method is 1 times less than that of the local map matching method. Build a two-level grid and store the property value for the whole Shenzhen urban area. The storage space is about 8MB, which can be read into memory at one time. The off-line calculation time of the whole road network in Shenzhen takes about 32 hours, but once the off-line calculation is completed, it can serve real-time matching for a long time.
【作者单位】：华北水利水电大学资源与环境学院;南京师范大学虚拟地理环境教育部重点实验室;信息工程大学地理空间信息学院;
【基金】：国家自然科学青年基金(41101373,41201467,41001313) 河南省高校科技创新团队支持计划资助(13IRTSTHN023) 南京师范大学虚拟地理环境教育部重点实验室开放基金资助
【分类号】：U495;TP391.41

【相似文献】