一种基于并行计算技术提高测井数据处理速度的方法  

A Method of Improving Processing Speed of Image Logging Data Based on Parallel Computing Technology

作　　者：何宗斌[1] 张宫[1] 樊鹤[1] 

机构地区：[1]油气资源与勘探技术教育部重点实验室(长江大学),湖北荆州434023

出　　处：《石油天然气学报》2012年第7期76-79,6-7,共4页Journal of Oil and Gas Technology

基　　金：国家科技重大专项(2011ZX05005-06)

摘　　要：基于多核处理器普及的背景,针对成像测井数据处理速度慢的问题,介绍了利用并行计算技术提高成像测井数据处理速度的方法。首次将并行计算技术应用于成像测井数据处理中,在Windows平台下编写了核磁数据反演的并行计算方法以及对应的串行计算方法,并在多个不同处理器的微机上对2种方法进行了对比。试验结果表明,在多核心处理器计算机上,并行计算能够充分利用CPU的效率,显著提高了数据处理的速度。同时,该方法的实现也为其他耗时型石油勘探数据处理提供了一种提高处理速度的新思路。Base on the current popularization of multicore processor,in allusion to the problem of slow speed of image logging data process,the parallel computing technology was introduced to improve the image log data processing speed.It was the first time to use the parallel computing technology in the image logging data processing.A parallel computing method of nuclear magnetic data inversion and corresponding serial calculating method were developed on the window platform,and they were compared on some microcomputers with different processors.The results show that on the computer with multi-core processors,parallel computing could make full use of CPU for obviously improving the data processing speed.Meanwhile,it provides a new idea for other consumptive petroleum exploration data processing.

关 键 词：并行计算 数据处理 多核 多线程 

分 类 号：P631.84[天文地球—地质矿产勘探]