融合注意力机制的二维卷积神经网络测井曲线重构方法

doi:10.13810/j.cnki.issn.1000-7210.2023.05.001

Abstract
Figure/Table
References
Related Citation (15)

Download: PDF (9617 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract The density and acoustic time-difference logging curves are two important curves that bridge seismic and rock physics, and they are the only logging curves that can provide reliable full-band elastic information for well-constrained seismic inversion. However, in practical applications, factors such as borehole collapse and instrument failures often result in distortion or missing data in density and acoustic time-difference logging. Moreover, existing empirical model methods, multivariate fitting methods, and rock physics modeling methods have limitations in accurately reconstructing target curves, especially when dealing with the simultaneous reconstruction of both curves. Therefore, this paper proposes to integrate the attention mechanism into the two-dimensional (2D) convolutional neural network to enhance the ability of the deep neural network (DNN) to capture the autocorrelation and cross-correlation features of the logging curves, so as to improve the accuracy of the DNN to reconstruct the acoustic and density logging curves. The ultra-deep tight sandstone in Junggar Basin is taken as the research object. Firstly, the relationship between the autocorrelation and cross-correlation features of logging curves and the weight distribution of the attention layer is analyzed. A comparative analysis is then conducted to assess the prediction accuracy of the proposed network against gated recurrent units (GRU) and 2D convolutional neural networks. The structural parameters of the proposed network are also optimized. Finally, the correction and reconstruction of the target curves are validated using synthetic seismic records, demonstrating the high prediction accuracy of the proposed network.

Key words： curve reconstruction attention mechanism 2D convolutional neural network deep learning acoustic logging density logging

Received: 03 January 2023

PACS:

P631

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	ZHAI Xiaoyan
	GAO Gang
	LI Yonggen
	CHEN Dong
	GUI Zhixian
	WANG Zhizhen

Cite this article:

ZHAI Xiaoyan,GAO Gang,LI Yonggen等. Reconstruction method of logging curves by 2D convolutional neural network integrating attention mechanism[J]. OGP, 2023, 58(5): 1031-1041.

URL:

http://www.ogp-cn.com/EN/10.13810/j.cnki.issn.1000-7210.2023.05.001 OR http://www.ogp-cn.com/EN/Y2023/V58/I5/1031

[1]	廖茂辉.多元回归方法校正扩径对密度曲线声波曲线的影响[J].物探与化探, 2014, 38(1):174-179, 184.LIAO Maohui.The application of multiple regression method to the calibration of the influence of hole enlargement on density and acoustic logs[J].Geophysical and Geochemical Exploration, 2014, 38(1):174-179, 184.
[2]	高刚, 杨亚华, 赵彬, 等.砂岩敏感识别因子的建立及直接提取方法[J].石油地球物理勘探, 2019, 54(6):1329-1338, 1347.GAO Gang, YANG Yahua, ZHAO Bin, et al.A method for establishing and directly extracting sensitive identification factors of unconsolidated sandstones[J].Oil Geophysical Prospecting, 2019, 54(6):1329-1338, 1347.
[3]	CASTAGNA J P, BATZLE M L, EASTWOOD R I.Relationship between compressional-wave and shear-wave velocities in clastic rocks[J].Geophysics, 1985, 50(5):571-581.
[4]	HAN D, NUR A, MORGAN D.Effects of porosity and clay content on wave velocities in sandstones[J].Geophysics, 1986, 51(11):2093-2107.
[5]	黄小平, 高玉春.对 Garden 公式应用条件的探讨[J].新疆石油地质, 2000, 21(3):233-234.HUANG Xiaoping, GAO Yuchun.An approach to application considerations of Garden's relation[J].Xinjiang Petroleum Geology, 2000, 21(3):233-234.
[6]	陈钢花, 王永刚.Faust 公式在声波曲线重构中的应用[J].勘探地球物理进展, 2005, 28(2):125-128.CHEN Ganghua, WANG Yonggang.Application of Faust equation in rebuilding of acoustic curve[J].Progress in Exploration Geophysics, 2005, 28(2):125-128.
[7]	熊冉, 高亮, 杨姣, 等.曲线重构反演在储层横向预测中的应用[J].西南石油大学学报(自然科学版), 2012, 34(1):83-89.XIONG Ran, GAO Liang, YANG Jiao, et al.Application of curve recomposition inversion in lateral reservoir prediction[J].Journal of Southwest Petroleum University (Science & Technology Edition), 2012, 34(1):83-89.
[8]	王俊瑞, 梁力文, 邓强, 等.基于多元回归模型重构测井曲线的方法研究及应用[J].岩性油气藏, 2016, 28(3):113-120.WANG Junrui, LIANG Liwen, DENG Qiang, et al.Research and application of log reconstruction based on multiple regression model[J].Lithologic Reservoirs, 2016, 28(3):113-120.
[9]	YANG Y, YIN X, GAO G, et al.Shear-wave velocity estimation for calciferous sandy shale formation[J].Journal of Geophysics and Engineering, 2019, 16(1):105-115.
[10]	桂俊川, 马天寿, 陈平.横观各向同性页岩岩石物理模型建立——以龙马溪组页岩为例[J].地球物理学报, 2020, 63(11):4188-4204.GUI Junchuan, MA Tianshou, CHEN Ping.Rock physics modeling of transversely isotropic shale:an example of the Longmaxi formation in the Sichuan basin[J].Chinese Journal of Geophysics, 2020, 63(11):4188-4204.
[11]	ZHAO W, CHELLAPPA R, PHILLIPS P J, et al.Face recognition:A literature survey[J].ACM Computing Surveys, 2003, 35(4):399-458.
[12]	YOUNG T, HAZARIKA D, PORIA S, et al.Recent trends in deep learning based natural language proce-ssing[J].IEEE Computational Intelligence Magazine, 2018, 13(3):55-75.
[13]	张家臣, 邓金根, 谭强, 等.基于XGBoost的测井曲线重构方法[J].石油地球物理勘探, 2022, 57(3):697-705.ZHANG Jiachen, DENG Jingen, TAN Qiang, et al.Reconstruction of well logs based on XGBoost[J].Oil Geophysical Prospecting, 2022, 57(3):697-705.
[14]	张益明, 张繁昌, 丁继才, 等.基于混合深度学习网络的致密砂岩甜点预测[J].石油物探, 2021, 60(6):995-1002.ZHANG Yiming, ZHANG Fanchang, DING Jicai, et al.Sweet spot prediction in tight sand reservoirs by a hybrid deep-learning net-work[J].Geophysical Prospec-ting for Petroleum, 2021, 60(6):995-1002
[15]	CHEN T F, GAO G, LIU H J, et al.Rock brittleness index inversion method with constraints of seismic and well logs via a CNN-GRU fusion network based on the spatiotemporal attention mechanism[J].Geoenergy Science and Engineering, 2023, 225:211646.
[16]	周伟, 赵海航, 蒋云凤, 等.基于串级双向长短时记忆神经网络的测井数据重构[J].石油地球物理勘探, 2022, 57(6):1473-1480.ZHOU Wei, ZHAO Haihang, JIANG Yunfeng, et al.Logging data reconstruction based on cascade bidirectional long short term memory neural network[J].Oil Geophysical Prospecting, 2022, 57(6):1473-1480.
[17]	HOCHREITER S, SCHMIDHUBER J.Long short-term memory[J].Neural Computation, 1997, 9(8):1735-1780.
[18]	CHEN T, GAO G, WANG P, et al.Prediction of shear wave velocity based on a hybrid network of two-dimensional convolutional neural network and gated recurrent unit[J].Geofluids, 2022, 14:9974157.
[19]	HAN J, LU C, CAO Z, et al.Integration of deep neural networks and ensemble learning machines for missing well logs estimation[J].Flow Measurement and Instrumentation, 2020, 73:101748.
[20]	DABRE R, CHU C, KUNCHUKUTTAN A.A survey of multilingual neural machine translation[J].ACM Computing Surveys, 2020, 53(5):1-38.
[21]	赵兵, 王增平, 纪维佳, 等.基于注意力机制的CNN-GRU短期电力负荷预测方法[J].电网技术, 2019, 43(12):4370-4376.ZHAO Bing, WANG Zengping, JI Weijia, et al.A short-term power load forecasting method based on attention mechanism of CNN-GRU[J].Power System Technology, 2019, 43(12):4370-4376.
[22]	王泽峰, 许辉群, 杨梦琼, 等.应用时域卷积神经网络的地震波阻抗反演方法[J].石油地球物理勘探, 2022, 57(2):279-286.WANG Zefeng, XU Huiqun, YANG Mengqiong, et al.Seismic impedance inversion method based on temporal convolutional neural network[J].Oil Geophysical Prospecting, 2022, 57(2):279-286.
[23]	常德宽, 雍学善, 王一惠, 等.基于深度卷积神经网络的地震数据断层识别方法[J].石油地球物理勘探, 2021, 56(1):1-8.CHANG Dekuan, YONG Xueshan, WANG Yihui, et al.Seismic fault interpretation based on deep convolutional neural networks[J].Oil Geophysical Prospecting, 2021, 56(1):1-8.
[24]	KAVIANPOUR P, KAVIANPOUR M, JAHANI E, et al.A CNN-BILSTM model with attention mechanism for earthquake prediction[J].The Journal of Supercomputing, 2023, doi:10.1007/s11227-023-05369-y.
[25]	BAI T, PEJMAN T.Attention-based LSTM-FCN for earthquake detection and location[J].Geophysical Journal International, 2021, 228(3):1568-1576.
[26]	马国庆, 王泽坤, 李丽丽.基于自注意力机制深度学习的重磁数据网格化和滤波方法[J].石油地球物理勘探, 2022, 57(1):34-42.MA Guoqing, WANG Zekun, LI Lili.Gridding and filtering method of gravity and magnetic data based on self attention deep learning[J].Oil Geophysical Prospecting, 2022, 57(1):34-42.