针对轨迹大数据的高效点-轨迹k近邻(point to trajectory k nearest neighbor, P2T_kNN)查询处理需求,提出了一种融合时空剖分和轨迹分段的轨迹组织方法,其核心思想是在对轨迹作时间剖分的基础上,利用离散全球网格系统(discrete global grid system, DGGS)在空间上进行再次剖分,从而利用两次剖分得到的时空单元编码来索引落入其中的轨迹片段。在此基础上利用分布式列式存储技术设计了面向轨迹大数据的P2T_kNN查询处理框架,提出了一种顾及轨迹数据空间分布的自适应空间单元搜索算法,即通过分析轨迹数据在给定时间约束下的空间分异特征,动态调整空间单元的搜索步长,从而提升了轨迹稀疏区域的处理效率。针对亿级轨迹的实验结果表明,该方法适用于轨迹大数据的P2T_kNN查询处理,在轨迹稠密与稀疏区域的平均查询响应时间均小于1 s。 相似文献
By using the observed monthly mean data over 160 stations of China and NCAR/NCEP reanalysis data, the generalized equilibrium feedback assessment(GEFA) method, combined with the methods of EOF analysis, correlation and composite analysis, is used to explore the influence of different SST modes on a wintertime air temperature pattern in which it is cold in the northeast and warm in the southwest in China. The results show that the 2009/2010 winter air temperature oscillation mode between the northern and southern part of China is closely related to the corresponding sea surface temperature anomalies(SSTA) and its associated atmospheric circulation anomalies. Exhibiting warming in Northeast China and cooling in Southwest China, the mode is significantly forced by the El Nio mode and the North Atlantic SSTA mode, which have three poles. Under the influence of SSTA modes, the surface northerly flow transported cold air to North and Northeast China, resulting in low temperatures in the regions. Meanwhile, the mid-latitude westerlies intensify and the polar cold air stays in high latitudes and cannot affect the Southwest China, resulting in the warming there. 相似文献