摘要
近年来各省级地震台网SEED文件数据量急增。在数据处理过程中,利用原有的串行解压缩算法批量解压缩地震波形数据时存在操作繁琐、耗时较长的问题。本文引入了MapReduce并行编程模型,根据该编程模型思想结合原有串行解压缩算法,提出了一种并行解压缩地震波形数据的算法,并给出了算法的设计与实现。本文从正确性、运行效率以及可扩展性三个方面进行了对比实验,验证了使用并行算法解压缩数据的效率较高,并且能够一次实现批量地震波形数据的解压缩,且操作简单。
In recent years, the number of SEED files was growing rapidly. In data processing, original algorithm of decompression batch seismic waveform data operated complicatedly and cost much time. In this paper, MapReduce programming model was introduced and a new parallel algorithm based on the thoughts of programming model and original decompression algorithm was presented. Also the design and implementation of this algorithm were given. Comparative experiments were carried out in terms of correctness, efficiency and extensibility. The results showed that the original algorithm spent more time compared to parallel algorithm which implementing decompression rapidly for a large number of seismic waveform data files. Using this method can decompress bulk of seismic waveform data and operate easily.
出处
《震灾防御技术》
CSCD
北大核心
2015年第2期344-352,共9页
Technology for Earthquake Disaster Prevention
基金
河北省重点地区壳幔结构及地震监测预报关键技术研究(13275407D)
河北省教育厅自然科学研究项目(QN20131141)
河北师范大学应用开发基金项目(L2012K01)联合资助
关键词
地震波形数据
解压缩
并行
Seismic waveform data
Decompress
Parallel
MapReduce