基于小波降噪与支持向量机的岩性预测方法  

Lithologic prediction method based on wavelet denoising and support vector machine

作　　者：程实 阎续 杨洁 CHENG Shi;YAN Xu;YANG Jie(Shale Gas Production Technology Service Company,Sinopec Oilfield Service Jianghan Corporation,Wuhan,Hubei 430042,China)

机构地区：[1]中石化江汉石油工程有限公司页岩气开采技术服务公司,湖北武汉430042

出　　处：《江汉石油职工大学学报》2022年第6期40-43,共4页Journal of Jianghan Petroleum University of Staff and Workers

摘　　要：随着机器学习和人工智能技术的发展,复杂地层岩性的自动估计已成为石油工程中最关键的要求之一。通过对正安两口水平井测井数据进行小波降噪处理,提升数据信噪比。同时利用声学(AC)、补偿中子(CNL)、密度(DEN)、伽马(GR)、光电(PE)、铀含量(U)和钍含量(TH)测井数据作为训练和测试样本,运用支持向量机(SVM)建立岩相识别模型。岩性识别准确度可达95.2%,相较于未采用降噪技术的人工智能模型相比,预测准确度提升近5个百分点。With the development of machine learning and artificial intelligence technology,the automatic evaluation of complex formation lithologies has become one of the most critical requirements in petroleum engineering.The signal-to-noise ratio(SNR)of well logging data in two horizontal wells in Zhengan area was improved by wavelet denoising processing.At the same time,the logging data such as acoustics(AC),compensated neutron(CNL),density(DEN),gamma(GR),photoelectric(PE),uranium content(U)and thorium content(TH)were used as training and testing samples,and the support vector machine(SVM)was used to establish the lithofacies identification model.The lithologic identification accuracy can reach 95.2%,which is nearly 5%higher than that of the artificial intelligence model without using the denoising technique.

关 键 词：支持向量机 小波降噪 岩性预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] P31[自动化与计算机技术—计算机科学与技术]