面向深度学习的Las格式测井数据转换器设计  

Design of Las Format Logging Data Converter for Deep Learning

作　　者：强伟帆 潘懋[1] 刘庆彬 郭艳军[1] 侯贵廷[1] QIANG Wei-fan;PAN Mao;LIU Qing-bin;GUO Yan-jun;HOU Gui-ting(The Key Laboratory of Orogenic Belts and Crustal Evolution(MOE),znstitute of Oil&Gas,School of Earth and Space Science,Peking University,Beijing 100871,China)

机构地区：[1]北京大学地球与空间科学学院造山带与地壳演化教育部重点实验室,石油与天然气研究中心,北京100871

出　　处：《科学技术与工程》2021年第1期248-253,共6页Science Technology and Engineering

基　　金：国家科技重大专项(2017ZX0513-002)。

摘　　要：深度学习在地质学领域有了越来越多的应用。在测井领域,可以利用深度学习方法更好地进行岩性分析、断层和裂隙识别等。测井数据最常用的数据格式是Las格式,而目前深度学习应用程序包不能直接对Las文件进行读取,因此存在数据间格式转换问题,需要将测井Las文件转化为能直接被程序包读取的数据格式。借助Visual Studio 2012平台,运用C#语言设计开发了一个LastoCSV格式转换软件,该软件能够对Las测井文件进行读取、分析、存储,然后转换为能被深度学习应用程序包直接读取的CSV格式文件,并且可对无效数据进行剔除处理。实验结果表明,该软件能快速准确地对数据进行处理并转换成CSV格式文件。Deep learning is increasingly widely applied in the field of geology.Regarding well logging,deep learning methods can be better used in lithological analysis,and fault and fracture identification.The most commonly used data format for logging data is the Las format.However,Las files cannot be read by deep learning programs because of format conversion problem among different types of data.It is necessary to convert the logging Las files into the data formats that can be read directly by the program.Therefore,the Visual Studio 2012 platform was used to design and develop a LastoCSV format conversion software using C# language.The software can read,analyze and store Las logging files,and then convert them into CSV format that can be directly read by deep learning application packages and eliminate invalid data.Experimental results show that the data can be converted and processed quickly and accurately.

关 键 词：测井曲线 深度学习 格式转化 C#语言 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]