硕士论文空间移动对象的最近邻查询方法研究移动对象连续k近邻查询.docVIP

Y 911771 分类号（中图法） TP3111 UPC (DDC) 005 密级 内部 . 论文作者姓名 吴林燕 学号0330703008单位河海大学 . 论文中文题名 空间移动对象的最近部查询方法研究 . 论文中文副题名 移动对象连续k近邻查询 . 论文英文题名 Research on Nearest Neighbor Query for Spatial Moving Objects . 论文英文副题名 Continuous k-Nearest Neighbor Query for Moving Objects 论文语种汉语论文摘要语种汉、英论文页数46论文字数3.3 (万） 论文主题词地理信息系统、连续k近邻查询、空间索引、R-tree、TPR-tree . 申请学位级别 工学硕士 专业名称计算机软件与理论 研究方向 空间数据库 指导教师姓名 朱跃龙教授 导师单位 河海大学计信学院 冯钩副教授导师单位 河海大学计信学院 论文答辩日期 2006年6月8曰Abstract With the rapid development of GPS and wireless communication, it is easy to trace and record the positions of moving objects. The research on nearest neighbor queries， especially the nearest neighbor queries of moving query and target objects, is arousing more and more interests and attentions. According to the track of moving objects, there is two different kinds of nearest neighbor queries of moving query and moving target objects. One of them, the continuous k-nearest neighbor queries without track is more popular in application. The traditional algorithm of continuous k-nearest neighbor queries is a series of static k-nearest neighbor queries, using snapshot in TPR-tree. This algorithm has a problem of repeating computation. It has to re-search all the k neighbors when query every time. This thesis improves the traditional algorithm of k-nearest neighbor queries on moving objects. Based on the truth that the results of every continuous two queries are related to each other when the interval of continuous queries is in some region, we present a candidate set to replace the whole target set. In this way, we can reduce the search region and enhance the query efficiency. Meanwhile, the candidate set can be diminished again by judging the trend of moving. Finally, we compare the improved algorithm to the traditional algorithm. The experiment proves the stability and efficiency of the improved algorithm is better than the traditional algorithm. Keywords: Geographical Information System，continuous k-nearest neighbor query， spatial