[1]杜睿山,李宏杰,孟令东.基于 CNN-BiLSTM-AM 的储层岩石脆性指数预测[J].计算机技术与发展,2023,33(10):28-34.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 005]
　DU Rui-shan,LI Hong-jie,MENG Ling-dong.Prediction of Reservoir Rock Brittleness Index Based on CNN-BiLSTM-AM[J].,2023,33(10):28-34.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 005]
点击复制

基于 CNN-BiLSTM-AM 的储层岩石脆性指数预测()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
33
期数:
2023年10期
页码:
28-34
栏目:
大数据与云计算
出版日期:
2023-10-10

文章信息/Info

Title:
Prediction of Reservoir Rock Brittleness Index Based on CNN-BiLSTM-AM
文章编号:
1673-629X(2023)10-0028-07
作者:
杜睿山1; 2 ; 李宏杰1 ; 孟令东2
1. 东北石油大学 计算机与信息技术学院,黑龙江 大庆 163318;
2. 油气藏及地下储库完整性评价黑龙江省重点实验室,黑龙江 大庆 163318
Author(s):
DU Rui-shan1; 2 ; LI Hong-jie1 ; MENG Ling-dong2
1. School of Computer and Information Technology,Northeast Petroleum University,Daqing 163318,China;
2. Key Laboratory of Oil and Gas Reservoir and Underground Gas Storage Integrity Evaluation,Daqing 163318,China
关键词:
测井曲线; 脆性指数; 深度学习; Pearson 系数; BiLSTM; 一维卷积; 注意力机制
Keywords:
logging curve; brittleness index; deep learning; Pearson coefficient; BiLSTM; one dimensional convolution; attention mechanism
分类号:
TP391. 1
DOI:
10. 3969 / j. issn. 1673-629X. 2023. 10. 005
摘要:
脆性指数是储层岩石的重要地质力学性质之一,但对于脆性指数至今为止都没有一个明确的定义,许多学者提出了不同的方法来测量该参数,一些方法如矿物分析等成本较高,然而机器学习和深度学习可以有效融合多元数据,充分利用数据去挖掘自变量与因变量之间的关系,且成本较低。 因此,该文使用深度学习方法来构建测井曲线数据与储层岩石脆性之间的关系模型。 因测井曲线是特殊的时序曲线,该文采用可以综合考虑过去和未来信息的 BiLSTM( 双向长短期记忆)模型,同时为了进一步的优化,在模型中添加 1DCNN( 一维卷积) 用来提取特征,且引入了 AM( 注意力机制)。 同时利用 Pearson 系数和 XGBoost( 极限梯度提升树) 进行分析,研究了各测井曲线对脆性的敏感性等级以及重要性程度,最终选取的输入参数有 AC( 声波时差) 、DEN(密度) 、CAL( 井径) 、GR( 伽马射线) 和 SP( 自然电位) 。 与其它机器学习方法相比,该方法的 MSE 和 RMSE 最小,分别为 0. 003 5 和 0. 059 16,表明 CNN-BiLSTM-AM 是一种预测精度更高、效果更好的方法。
Abstract:
Brittleness index is one of the important geomechanical properties of reservoir rocks, but there is no clear definition forbrittleness index so far. Many scholars had proposed different methods to measure this parameter. Some methods, such as mineralanalysis,had higher costs. However,machine learning and depth learning can effectively integrate multivariate data,and make full use ofthe data to mine the relationship between independent variables and dependent variables with lower costs. Therefore, we used depthlearning method to build the relationship model between logging curve data and reservoir rock brittleness. Because the logging curve wasa special time series curve, we adopted the BiLSTM ( Bi - directional Long Short - Term Memory ) model that can comprehensivelyconsider past and future information. At the same time,for further optimization,1DCNN ( One-Dimensional Convolution) was added tothe model to extract features, and AM ( Attention Mechanism) was introduced. At the same time, Pearson coefficient and XGBoost( eXtreme Gradient Boosting) were used for analysis,and the sensitivity level and importance of each logging curve to brittleness arestudied. The final selected input parameters were AC ( Acoustic moveout) ,DEN ( Density) ,CAL ( Caliper) ,GR ( Gamma Ray) and SP( Spontaneous Potential) . Compared with other machine learning methods, the proposed method has the smallest MSE and RMSE,0. 003 5 and 0. 059 16 respectively. It is showed that CNN-BiLSTM-AM is a method with higher prediction accuracy and better effect.

相似文献/References:

[1]尚福华[],柴艳领[],杜睿山[],等. 小波双线性插值算法在测井曲线相似中的应用[J].计算机技术与发展,2015,25(09):66.
　HANG Fu-hua[],CHAI Yan-ling[],DU Rui-shan[],et al. Application of Wavelet Bilinear Interpolation Algorithm in Well-logging Curve Similarity[J].,2015,25(10):66.
[2]柴艳领[],唐志涛[].Haar 小波变换在测井曲线相似性对比中的应用[J].计算机技术与发展,2016,26(01):119.
　CHAI Yan-ling[],TANG Zhi-tao[]. pplication of Haar Wavelet Transform in Well-logging Curve Similarity[J].,2016,26(10):119.
[3]尚福华,李金成*,原 野,等.基于改进 BP 神经网络的地层划分方法[J].计算机技术与发展,2020,30(09):148.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 027]
　SHANG Fu-hua,LI Jin-cheng*,YUAN Ye,et al.Stratigraphic Division Method Based on Improved BP Neural Network[J].,2020,30(10):148.[doi:10. 3969 / j. issn. 1673-629X. 2020. 09. 027]
[4]冯向东,张玉琴,韩红伟,等.基于自组织神经网络在油气分层中的研究[J].计算机技术与发展,2021,31(02):44.[doi:10. 3969 / j. issn. 1673-629X. 2021. 02. 008]
　FENG Xiang-dong,ZHANG Yu-qin,HAN Hong-wei,et al.Research on Oil and Gas Stratification Based on Self-organizing Neural Network[J].,2021,31(10):44.[doi:10. 3969 / j. issn. 1673-629X. 2021. 02. 008]
[5]富 宇,李 滕,郭晓萍.基于支持向量机对优势渗流通道识别的研究[J].计算机技术与发展,2021,31(08):182.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 031]
　FU Yu,LI Teng,GUO Xiao-ping.Research on Identification of Dominant Seepage ChannelBased on Support Vector Machine[J].,2021,31(10):182.[doi:10. 3969 / j. issn. 1673-629X. 2021. 08. 031]
[6]肖 红,张瑶瑶*,张福禄.改进的卷积神经网络及在地层识别中的应用[J].计算机技术与发展,2021,31(09):167.[doi:10. 3969 / j. issn. 1673-629X. 2021. 09. 028]
　XIAO Hong,ZHANG Yao-yao*,ZHANG Fu-lu.Improved Convolution Neural Network and Its Application of Stratigraphic Identification[J].,2021,31(10):167.[doi:10. 3969 / j. issn. 1673-629X. 2021. 09. 028]
[7]张可佳,方佳佳,高楷程,等.基于神经网络的岩性物性参数计算方法研究[J].计算机技术与发展,2022,32(06):186.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 031]
　ZHANG Ke-jia,FANG Jia-jia*,GAO Kai-cheng,et al.Study on Calculation Method of Lithologic Physical Property Parameters Based on Neural Network[J].,2022,32(10):186.[doi:10. 3969 / j. issn. 1673-629X. 2022. 06. 031]
[8]曹茂俊,崔欣锋.基于一维卷积神经网络的地层智能识别方法[J].计算机技术与发展,2023,33(09):133.[doi:10. 3969 / j. issn. 1673-629X. 2023. 09. 020]
　CAO Mao-jun,CUI Xin-feng.Intelligent Stratigraphic Recognition Method Based on 1DCNN[J].,2023,33(10):133.[doi:10. 3969 / j. issn. 1673-629X. 2023. 09. 020]

更新日期/Last Update: 2023-10-10