基于双树双索引结构的移动查询方法  

Moving Query Method Based on Double R-Tree and Double Indexing

作　　者：梁野[1] 龚雪晶[1] 慈林林[1] 苏斐[2] 

机构地区：[1]北京理工大学计算机科学与技术学院,北京100081 [2]中国石油信息技术服务中心,北京100724

出　　处：《北京理工大学学报》2008年第11期993-997,共5页Transactions of Beijing Institute of Technology

基　　金：国家部委预研项目(41316040104)

摘　　要：为实现对有限范围内海量移动对象的有效索引,构建通用的移动查询解决方案,针对移动对象在道路网格中的运动特点,提出了预测实时运动速度的速度积累模型和预测未来聚集位置的基于双树双索引结构的移动对象查询方法.双树双索引结构利用网格划分思想构建空间分割树,实现对现有GG TPR-tree查询结构的拓展,并结合GG TPR-tree索引及建立于内存中的Hash索引以满足各种类型的移动查询请求.仿真实验表明,在回答受限范围内海量移动对象的确定性查询和统计性查询时,与传统方法相比,双树双索引结构在查询结果准确率方面有明显的改善.In order to effectively index a huge amount of moving objects within a limited area, and to put forward a universal solution to response moving query, according to the mobile characteristic of moving objects in the road grid, speed accumulation model which can predict the current speed and the moving query method based on double R-tree and double indexing （DTDI） structure that can predict where the moving objects will be in the future are proposed. The DTDI structure builds up the space dividing tree based on grid to expand the already existing GG TPR- tree query structure. The space dividing tree, the GG TPR-tree, the GG TPR-tree indexing, and the Hash indexing in the memory together make up the DTDI structure to meet the requirements for all kinds of moving queries. Experimental results showed that the accuracy for the DTDI structure is better than other structures in querying a great capacity of moving objects within a limited area.

关 键 词：移动查询 网格 GG TPR树 空间分割树 双树双索引 

分 类 号：TP393.103[自动化与计算机技术—计算机应用技术]