• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2006, Vol. 28 ›› Issue (10): 74-77.

• 论文 • 上一篇    下一篇

基于时空连续性的运动对象最近邻居查询

潘鹏 卢炎生   

  • 出版日期:2006-10-01 发布日期:2010-05-20

  • Online:2006-10-01 Published:2010-05-20

摘要:

时空数据库中的运动对象最近邻居查询是NN Queries中的新问题,基于TPR-TREE索引结构的TP NN Queries算法能较好地处理对象的时态特性,但会多次查询同一对象。本文 利用运动对象的时空连续性对TP NN Queries算法进行改进,通过一次查询TPR-TREE索引获取所有候选NN对象与查询对象的距离变化曲线,进而得到NN对象集,减少了查询及 时空运算的次数。本文最后给出了实验分析。

关键词: 时空数据库 最近邻居 TPR-TREE 时空连续性

Abstract:

In spatio-temporal databases, the moving objects" nearest neighbor query is a new problem ot nearest neighbor queries(NN Queries). The current Tp NN Queries method which is based on the index structure of TPR-TREE can deal with moving objects, but it has the shortcomings of repeatedly querying somemoving objects. Based on the theory of moving objects' spatio-temporal continuity, we propose an improved method(TP NN Queries) that searches the TP PR-TREE once to get the curve of distance changes between the query object and the candidate nearest neighbor set, and then we get the NN set. Our method cuts down the query and calculation times. Experimental results show that our method makes a practical improvement.

Key words: (spatio-temporal database, NN;TPR-TREE, spatio-temporal continuity)