基于LSTM循环神经网络的岩性识别方法

doi:10.12108/yxyqc.20210312

岩性油气藏 ›› 2021, Vol. 33 ›› Issue (3): 120–128.doi: 10.12108/yxyqc.20210312

基于LSTM循环神经网络的岩性识别方法

武中原¹, 张欣², 张春雷³, 王海英¹

1. 中国地质大学(北京)数理学院, 北京 100083;
2. 北京师范大学统计学院, 北京 100875;
3. 北京中地润德石油科技有限公司, 北京 100083

收稿日期:2020-07-22 修回日期:2020-09-06 发布日期:2021-06-03
第一作者:武中原（1996—），男，中国地质大学（北京）在读硕士研究生，研究方向为统计学习、机器学习。地址：（100083）北京市海淀区学院路29号。Email：937031303@qq.com。
基金资助:
国家科技重大专项“鄂尔多斯盆地大型岩性地层油气藏勘探开发示范工程”（编号：2016ZX05050）资助

Lithology identification based on LSTM recurrent neural network

WU Zhongyuan¹, ZHANG Xin², ZHANG Chunlei³, WANG Haiying¹

1. School of Science, China University of Geosciences(Beijing), Beijing 100083, China;
2. School of Statistics, Beijing Normal University, Beijing 100875, China;
3. Beijing Zhongdirunde Petroleum Technology Co., Ltd., Beijing 100083, China

Received:2020-07-22 Revised:2020-09-06 Published:2021-06-03

摘要/Abstract

摘要： 针对复杂碳酸盐岩储层岩石组分复杂、岩性多样，常规测井岩性识别方法受限等问题，提出利用长短期记忆神经网络（LSTM）提高岩性识别效果的方法，并结合实际数据进行验证和应用效果分析。考虑到常规机器学习方法在岩性识别中无法充分利用沉积岩石在深度域序列上的潜在信息，从而基于LSTM方法构建了能够提取和学习岩性沉积序列特征的岩性识别手段。以苏里格气田苏东地区下古生界碳酸盐岩储层为例，通过敏感性分析选取自然伽马、光电吸收截面指数、密度、声波时差、补偿中子和电阻率等6种测井参数，构建了基于LSTM的岩性识别模型。结果表明，与朴素贝叶斯，KNN，决策树，SVM和HMM等传统方法相比，LSTM的岩性识别准确率提升幅度介于1.40%~12.25%。高精度的LSTM岩性识别模型为复杂碳酸盐岩储层的表征和评价提供了数据基础。

关键词: 长短期记忆神经网络, 岩性识别, 碳酸盐岩储层, 机器学习

Abstract: A lithology recognition method by long-short-term memory neural network（LSTM）was proposed for complex carbonate reservoirs with complex composition and diverse lithology，to overcome obstacles troubling traditional identification，and effective results were showed with a case from gas field. Due to the inadequate ability of general machine learning methods in extracting the characteristics of sedimentary sequence，the LSTM method was introduced into the improvement for lithology identification. Taking the Lower Paleozoic carbonate reservoir in eastern block of Sulige gas field as an example，six sensitive parameters were selected to construct a lithology identification model based on LSTM，such as GR，Pe，DEN，R_LLD， AC and CNL. The results show that lithology identification accuracy based on LSTM increases by 1.40%-12.25% above traditional models（Naive Bayes，KNN，Decision Tree，SVM and HMM），and can provide more reliable support for the characterization and evaluation of complex carbonate reservoirs.

Key words: long-short-term memory neural network, lithology identification, carbonate reservoir, machine learning

中图分类号:

P618.13

武中原, 张欣, 张春雷, 王海英. 基于LSTM循环神经网络的岩性识别方法[J]. 岩性油气藏, 2021, 33(3): 120–128.

WU Zhongyuan, ZHANG Xin, ZHANG Chunlei, WANG Haiying. Lithology identification based on LSTM recurrent neural network[J]. Lithologic Reservoirs, 2021, 33(3): 120–128.

参考文献

[1] 袁照威,段正军,张春雨,等.基于马尔科夫概率模型的碳酸盐岩储集层测井岩性解释.新疆石油地质,2017,38(1):96-102. YUAN Z W,DUAN Z J,ZHANG C Y,et al. Interpretation of logging lithology in carbonate reservoirs based on Markov Chain probability model. Xinjiang Petroleum Geology,2017,38(1):96-102.
[2] 成大伟,袁选俊,周川闽,等.测井岩性识别方法及应用:以鄂尔多斯盆地中西部长7油层组为例.中国石油勘探,2016,21(5):117-126. CHENG D W,YUAN X J,ZHOU C M,et al. Logging lithology identification methods and their application:A case study on Chang 7 member in central-western Ordos Basin,NW China. China Petroleum Exploration,2016,21(5):117-126.
[3] 王泽华,朱筱敏,孙中春,等.测井资料用于盆地中火成岩岩性识别及岩相划分:以准噶尔盆地为例.地学前缘,2015,22(3):254-268. WANG Z H,ZHU X M,SUN Z C,et al. Igneous lithology identification and lithofacies classification in the basin using logging data:Taking Junggar Basin as an example. Earth Science Frontiers,2015,22(3):254-268.
[4] 马峥,张春雷,高世臣.主成分分析与模糊识别在岩性识别中的应用.岩性油气藏,2017,29(5):127-133. MA Z,ZHANG C L,GAO S C. Lithology identification based on principal component analysis and fuzzy recognition. Lithologic Reservoirs,2017,29(5):127-133.
[5] 王振洲,张春雷,高世臣.利用决策树方法识别复杂碳酸盐岩岩性:以苏里格气田苏东41-33区块为例.油气地质与采收率,2017,24(6):25-33. WANG Z Z,ZHANG C L,GAO S C. Lithology identification of complex carbonate rocks based on decision tree method:An example from block Sudong 41-33 in Sulige gas field. Petroleum Geology and Recovery Efficiency,2017,24(6):25-33.
[6] 孙予舒,黄芸,梁婷,等.基于XGBoost算法的复杂碳酸盐岩岩性测井识别.岩性油气藏,2020,32(4):98-106. SUN Y S,HUANG Y,LIANG T,et al. Identification of complex carbonate lithology by logging based on XGBoost algorithm. Lithologic Reservoirs,2020,32(4):98-106.
[7] AL-ANAZI A,GATES I D. A support vector machine algorithm to classify lithofacies and model permeability in heterogeneous reservoirs. Engineering Geology,2010,114(3/4):267-277.
[8] 袁照威,陈龙,高世臣,等.基于马尔科夫-贝叶斯模拟算法的多地震属性沉积相建模方法:以苏里格气田苏10区块为例. 油气地质与采收率,2017,24(3):37-43. YUAN Z W,CHEN L,GAO S C,et al. A method of sedimentary facies modeling through integration of multi-seismic attributes based on Markov-Bayes model:An example from Su10 area in the north of Sulige gas field. Petroleum Geology and Recovery Efficiency,2017,24(3):37-43.
[9] 仲鸿儒,成育红,林孟雄,等.基于SOM和模糊识别的复杂碳酸盐岩岩性识别.岩性油气藏,2019,31(5):84-91. ZHONG H R,CHENG Y H,LIN M X,et al. Lithology identification of complex carbonate based on SOM and fuzzy recognition. Lithologic Reservoirs,2019,31(5):84-91.
[10] 刘跃杰,刘书强,马强,等. BP神经网络法在三塘湖盆地芦草沟组页岩岩相识别中的应用.岩性油气藏,2019,31(4):101-111. LIU Y J,LIU S Q,MA Q,et al. Application of BP neutral network method to identification of shale lithofacies of Lucaogou Formation in Santanghu Basin. Lithologic Reservoirs,2019,31(4):101-111.
[11] ELFEKI A,DEKKING M. A Markov Chain model for subsurface characterization:Theory and applications. Mathematical Geology,2001,33(5):569-589.
[12] LINDBERG D V,GRANA D. Petro-elastic log-facies classification using the expectation maximization algorithm and hidden markov models. Math Geosciences,2015,47(6):719-752.
[13] HOCHREITER S,SCHMIDHUBER J. Long short-term memory. Neural Computation,1997,9(8):1735-1780.
[14] 张东晓,陈云天,孟晋.基于循环神经网络的测井曲线生成方法. 石油勘探与开发,2018,45(4):598-607. ZHANG D X,CHEN Y T,MENG J. Synthetic well logs generation via recurrent neural networks. Petroleum Exploration and Development,2018,45(4):598-607.
[15] ZHANG J F,ZHU Y,ZHANG X P,et al. Developing a long short-term memory(LSTM)based model for predicting water table depth in agricultural areas. Journal of Hydrology,2018,6(561):918-929.
[16] BAO W,YUE J L,RAO Y L. A deep learning framework for financial time series using stacked autoencoders and long-short term memory. Plos One,2017,12(7):e0180944.
[17] SCHUSTER M,PALIWAL K K. Bidirectional recurrent neural networks. IEEE Transactions on Signal Processing,1997,45(11):2673-2681.
[18] BENGIO Y,SIMARD P,FRASCONI P. Learning long-term dependencies with Gradient Descent is difficult. IEEE Trans Neural Network,2002,5(2):157-166.
[19] GRAVES A,JAITLY N,Mohamed A R. Hybrid speech recognition with deep bidirectional LSTM. Automatic Speech Recognition and Understanding(ASRU),2013 IEEE Workshop on. IEEE,2013.
[20] 罗群,吴安彬,王井伶,等.中国北方页岩气成因类型、成气模式与勘探方向.岩性油气藏,2019,31(1):1-11. LUO Q,WU A B,WANG J L,et al. Genetic types,generation models,and exploration direction of shale gas in northern China. Lithologic Reservoirs,2019,31(1):1-11.
[21] 靳军,王剑,杨召,等. 准噶尔盆地克-百断裂带石炭系内幕储层测井岩性识别.岩性油气藏,2018,30(2):85-92. JIN J,WANG J,YANG Z,et al. Welling logging identification of Carboniferous volcanic inner buried-hill reservoirs in Ke-Bai fault zone in Junggar Basin. Lithologic Reservoirs,2018,30(2):85-92.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于LSTM循环神经网络的岩性识别方法

Lithology identification based on LSTM recurrent neural network

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 10

[1]	赵军, 李勇, 文晓峰, 徐文远, 焦世祥. 基于斑马算法优化支持向量回归机模型预测页岩地层压力[J]. 岩性油气藏, 2024, 36(6): 12-22.
[2]	张天择, 王红军, 张良杰, 张文起, 谢明贤, 雷明, 郭强, 张雪锐. 射线域弹性阻抗反演在阿姆河右岸碳酸盐岩气藏储层预测中的应用[J]. 岩性油气藏, 2024, 36(6): 56-65.
[3]	桂金咏, 李胜军, 高建虎, 刘炳杨, 郭欣. 基于特征变量扩展的含气饱和度随机森林预测方法[J]. 岩性油气藏, 2024, 36(2): 65-75.
[4]	陈叔阳, 何云峰, 王立鑫, 尚浩杰, 杨昕睿, 尹艳树. 塔里木盆地顺北1号断裂带奥陶系碳酸盐岩储层结构表征及三维地质建模[J]. 岩性油气藏, 2024, 36(2): 124-135.
[5]	杨午阳, 魏新建, 李海山. 智能物探技术的过去、现在与未来[J]. 岩性油气藏, 2024, 36(2): 170-188.
[6]	郭海峰, 肖坤叶, 程晓东, 杜业波, 杜旭东, 倪国辉, 李贤兵, 计然. 乍得Bongor盆地花岗岩潜山裂缝型储层有效渗透率计算方法[J]. 岩性油气藏, 2023, 35(6): 117-126.
[7]	唐昱哲, 柴辉, 王红军, 张良杰, 陈鹏羽, 张文起, 蒋凌志, 潘兴明. 中亚阿姆河右岸东部地区侏罗系盐下碳酸盐岩储层特征及预测新方法[J]. 岩性油气藏, 2023, 35(6): 147-158.
[8]	许鑫, 杨午阳, 张凯, 魏新建, 张向阳, 李海山. 三维初至波旅行时层析速度反演算法优化[J]. 岩性油气藏, 2023, 35(4): 79-89.
[9]	宋兴国, 陈石, 杨明慧, 谢舟, 康鹏飞, 李婷, 陈九洲, 彭梓俊. 塔里木盆地富满油田F_Ⅰ16断裂发育特征及其对油气分布的影响[J]. 岩性油气藏, 2023, 35(3): 99-109.
[10]	倪新锋, 沈安江, 乔占峰, 郑剑锋, 郑兴平, 杨钊. 塔里木盆地奥陶系缝洞型碳酸盐岩岩溶储层成因及勘探启示[J]. 岩性油气藏, 2023, 35(2): 144-158.
[11]	宋传真, 马翠玉. 塔河油田奥陶系缝洞型油藏油水流动规律[J]. 岩性油气藏, 2022, 34(4): 150-158.
[12]	叶涛, 牛成民, 王清斌, 高坤顺, 孙哲, 陈安清. 用“成分-结构”分类法识别古潜山变质岩岩性——以渤海海域太古界为例[J]. 岩性油气藏, 2021, 33(6): 156-164.
[13]	李树博, 郭旭光, 郑孟林, 王泽胜, 刘新龙. 准噶尔盆地东部西泉地区石炭系火山岩岩性识别[J]. 岩性油气藏, 2021, 33(1): 258-266.
[14]	孙予舒, 黄芸, 梁婷, 季汉成, 向鹏飞, 徐新蓉. 基于XGBoost算法的复杂碳酸盐岩岩性测井识别[J]. 岩性油气藏, 2020, 32(4): 98-106.
[15]	宋宣毅, 刘月田, 马晶, 王俊强, 孔祥明, 任兴南. 基于灰狼算法优化的支持向量机产能预测[J]. 岩性油气藏, 2020, 32(2): 134-140.