一种基于格网划分的浮动车数据自适应地图匹配方法研究  

Study on an Algorithm of Auto-Adaptive Map Matching for Floating Car Data Based on Grid Division

作　　者：陶华超[1] 马林兵[1] 魏慧丽[1] 周群[1] 

机构地区：[1]中山大学地理科学与规划学院,广东广州510275

出　　处：《地理与地理信息科学》2015年第3期22-25,37,共5页Geography and Geo-Information Science

基　　金："十二五"国家科技支撑项目(37000-41070400)

摘　　要：利用单一的匹配算法对区域内的浮动车数据进行地图匹配,会出现浮动车点匹配到邻近路段上的跳跃现象。该文将区域划分格网,遍历待匹配点所在格网及其8邻域格网,筛选出候选路段、结点集合;根据候选路段、结点数量特征,自主选择合适的算法,计算几何距离和匹配度指标以评价匹配结果,确保匹配准确性。通过广州的部分区域数据进行算法验证表明:基于合适步长的格网划分能够提高匹配效率;与单一的最近点匹配算法相比,自适应综合匹配算法能够较好地避免"点跳跃",提高匹配准确度。A new method is proposed in this paper that is an auto-adaptive approach to deal with FlED to match the map based on grid. It traverses the grids including the prepared-matching point and its eight neighboring grids after the computational domain is divided into suitable MX N grids, and then filters out candidate sections or nodes collection. According to the quantitative characteristics of candidate sections or nodes, it is auto enough to choose the appropriate algorithm to calculate. To ensure the matching accuracy, the geometric distance and matching degree are introduced to evaluate the matching result during the imple- mentation process. The experiment indicates that the matching efficiency will be affected by the step length of grid. And com- bined with matching degree, the auto-adaptive matching algorithm is better to avoid ＂jumps＂ and can increase matching accuracy comparing with the closest point matching algorithm. Discussions of the benefits and shortcomings associated with this method are provided, along with suggestion for future research.

关 键 词：浮动车数据 地图匹配 自适应 格网 匹配度 

分 类 号：P208[天文地球—地图制图学与地理信息工程] U12[天文地球—测绘科学与技术]