面向海量星表数据的高效的时序数据生成方法研究  

EFFICIENT TIME SERIES DATA GENERATION METHOD FOR MASSIVE STAR TABLE DATA

作　　者：熊聪聪[1] 付立艳 赵青[1] Xiong Congcong;Fu Liyan;Zhao Qing(College of Artificial Intelligence,Tianjin University of Science&Technology,Tianjin 300457,China)

机构地区：[1]天津科技大学人工智能学院,天津300457

出　　处：《计算机应用与软件》2021年第4期17-23,共7页Computer Applications and Software

基　　金：国家自然科学基金青年项目(11803022);天津市科委自然科学基金青年项目(18JCQNJC69800);天津市教委科研计划项目(2017KJ035)。

摘　　要：在时序天文学中,观测设备具有在时间轴上频繁采样的特点,这导致处理数据的海量性,应用传统科学计算方法生成时序数据通常耗时较长,效率不高。为了解决这个问题,提出一种高效时序数据生成方法(ETSDGM)。根据用户的时空请求实现时序数据的快速生成,重点设计对原始文件的数据访存优化和交叉证认计算过程中的证认计算优化。实验结果表明,在数据量较大的情况下,ETSDGM可以实现更好的性能改进。In time series astronomy,observation equipment has the characteristic of frequent sampling on the time axis,which leads to massive processing data.It usually takes a long time and has low efficiency to generate time series data using traditional scientific calculation methods.In order to solve this problem,an efficient time series data generation method(ETSDGM)is proposed.According to the user s space-time request,time series data can be generated quickly.The optimization of data access and storage of original files and the optimization of authentication calculation in the process of cross-authentication calculation were mainly designed.The experimental results show that ETSDGM can achieve better performance improvement.

关 键 词：时序数据 时空子集 访存优化 计算优化 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]