基于 MongoDB 的海量天文星表数据的快速时序重构研究
天文数据的爆发性增长,导致运用传统科学计算方法生成天文时序数据时效率不高,直接影响时域天文学的科学产出。为了解决这一问题,文章提出了减少距离计算的同源星表快速证认方法及基于 MongoDB 的应用方案,重点从原始数据的访存优化,证认计算速度的提高等方面寻求新的改进方案,以解决大规模天文星表的批量时序重构的效率问题。实验结果表明,与基于传统多波段交叉证认算法和关系型数据库的方法相比,该方法可以更有效地提高时序数据的生成效率,为时域天文学时代频繁采样望远镜大规模星表数据的时序重构和光变曲线的生成提供了新思路。
he explosive growth of astronomical data leads to low efficiency in the generation of astronomical time series data by traditional scientific calculation methods, which directly affects the scientific output of time domain astronomy. In this paper, we propose a fast method to authenticate the same catalog and a MongoDB-based application to reduce the distance computation. In order to solve the efficiency problem of batch time series reconstruction of large-scale astronomical catalogues, we focus on the optimization of original data access and the improvement of authentication computation speed. The experimental results show that this method can improve the efficiency of time series data generation more effectively than the traditional multi-band cross-validation algorithm and relational database method, it provides a new idea for the reconstruction of time series and the generation of light curves for the large-scale catalogue data of the time-domain astronomical time-frequency sampling telescope.
徐丹滢,赵 青,权文利,宋红壮
10.3969/j.issn.1000-8349.2022.02.10
天文学
MongoDB地理空间索引访存优化证认优化
徐丹滢,赵 青,权文利,宋红壮.基于 MongoDB 的海量天文星表数据的快速时序重构研究[EB/OL].(2023-06-07)[2025-08-10].https://chinaxiv.org/abs/202306.00384.点此复制
评论