考虑运动对象位置不确定性的位置管理策略  

Tactics of location management for moving objects with uncertainty

作　　者：陈基雄[1] 朱玺[1] 

机构地区：[1]华中科技大学计算机科学与技术学院,湖北武汉430074

出　　处：《华中科技大学学报（自然科学版）》2005年第11期22-25,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60203017)

摘　　要：以传统轨道模型为基础,观测构造运动对象历史信息统计样本并分析其特征,设计了基于样本统计的位置预测模型,更好地考虑了运动对象运动随机性的特点,提高了位置预测的准确性;然后在该模型的基础上设计了一种为管理运动对象位置不确定性而设定最佳阈值的方法,这种方法相对于传统的静态阈值策略减少了位置更新的信息代价开销;最后,采用一种基于时间和空间划分的Grid模型构造索引结构,给出了管理运动对象位置信息进行区域查询和kNN(k-Nearest Neighbor)查询的实现过程和算法,是一种进行运动对象位置相关查询的可行性方案.On the basis of the traditional trajectory model, the model for predicting the position was designed by means of observing object historical information, constructing statistic sample and analyzing its characteristics in accordance with the sample statistics. Owing to taking of the random of moving object into consideration, this model has improved the accuracy of location prediction. The optimum threshold method for managing uncertainty of moving objects was constructed based on this model, decreasing the cost of location updating by comparing the traditional static threshold policy. The region query of moving objects and the processing of kNN （k-Nearest Neighbor） query and its algorithm were provided by the index construction of the Grid model based on the division of time and space. The policy is feasible for querying the location of the moving objects.

关 键 词：位置查询 运动对象 不确定性管理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]