约束的移动对象最近邻轨迹查询

Nearest neighbor search for constrained moving objects trajectories

提出了一种新的查询概念,即约束的移动对象最近邻轨迹(constrained nearest neighbor search on mov-ing objects trajectories,简称CNNMOT)查询.针对两阶段算法分阶段处理最近邻查询和范围查询的缺陷,对传统空间最近邻算法进行了改进和扩展,使其能够合并到一步完成,并对传统的mindist定义进行了修改.实验结果表明:该改进算法能较好地克服两阶段算法的缺陷,并在数据量较大时仍保持较高的性能.

A new notion of query, i.e. constrained nearest neighbor search for moving objects trajectories （CNNMOT） was introduced. After the disadvantages of two phase algorithm, which processes nearest neighbor query and range query in step, were considered, the spatial nearest neighbor query algorithm was modified and extended to integrate into one step. The traditional notion of mindist was also revised. Experimental results show that the modified algorithm can overcome the shortcomings of two-tier algorithm and keep its higher performance when the data are numeruous.