Abstract：The gridding and filtering of gravity and magnetic data directly influence the result of data processing. This paper designs a more rational deep learning model to improve the accuracy of gridding and filtering. The gridding method based on self-attention deep learning is constructed, and the self-attention mechanism layer is utilized to process the two-dimensional position embeddings. In this way, the position embeddings vector is obtained with global and local information integrated. Then the position information and anomaly information are fused to output node anomaly, thus reducing the distortion. For the random and stripe noises of gravity and magnetic data, a convolutional neural network is first employed to classify noise. The stripe noise is filtered by self-attention convolutional neural network and the random noise by convolutional autoencoder to get high-quality basic data. Model experiment shows that the gridded structure of deep learning is closer to the real result than that of the traditional method. The proposed filtering method can remove various noises, providing more accurate basic data for the follo-wing inversion. The gridding and filtering method based on deep learning applied to practical magne-tic data achieves good results, proving that it has strong feasibility and practicability.
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. Oil Geophysical Prospecting, 2022, 57(1): 34-42.
HARDY R L.Multiquadric equations of topography and other irregular surfaces[J].Journal of Geophysical Research, 1971, 76(8):1905-1915.
[2]
BRIGGS I.Machine contouring using minimum curvature[J].Geophysics, 1974, 39(1):39-48.
[3]
SWAIN C J.A FORTRAN IV program for interpolating irregular spaced data using the difference equations for minimum curvature[J].Computers & Geosciences, 1976, 1(4):231-240.
[4]
SMITH W H F, WESSEL P.Gridding with continues curvature splines in tension[J].Geophysics, 1990, 55(3):293-305.
[5]
王万银,邱之云,刘金兰,等.位场数据处理中的最小曲率扩边和补空方法研究[J].地球物理学进展,2009, 24(4):1327-1338.WANG Wanyin, QIU Zhiyun, LIU Jinlan, et al.Research on the method of minimum curvature expanding edge and filling gap in potential field data processing[J].Progress in Geophysics, 2009, 24(4):1327-1338.
[6]
刘兆平,杨进,武炜.地球物理数据网格化方法的选取[J].物探与化探,2010,34(1):93-97.LIU Zhaoping, YANG Jin, WU Wei.Selection of geophysics data grid method[J].Geophysical and Geochemical Exploration, 2010, 34(1):93-97.
[7]
HANSEN R O.Interpretive gridding by anisotropic Krigging[J].Geophysics, 1993, 58(10):1491-1497.
[8]
孙洪泉.地质统计学及其应用[M].北京:中国矿业大学出版社, 1990.
[9]
CORDELL L.A scattered equivalent soured method for interpolation and gridding of potential field data in three dimensions[J].Geophysics, 2012, 77(4):P629-P636.
[10]
MALLET J L.Discrete smooth and interpolation in geometric modelling[J].Computer Aided Design, 1992, 24(4):178-191.
[11]
孟小红,侯建全,梁宏英,等.离散光滑插值方法在地球物理位场中的快速实现[J].物探与化探,2002,26(4):302-306.MENG Xiaohong, HOU Jianquan, LIANG Hongying, et al.The fast realization of discrete smooth interpolation in the interpolation of potential data[J].Geophysical and Geochemical Exploration, 2002, 26(4):302-306.
[12]
CLAERBOUT J F.Image Estimation by Example[D].Stanford University, California,2002.
[13]
李冰,刘洪,李幼铭.三维地震数据离散光滑插值的共轭梯度法[J].地球物理学报,2002, 45(5):691-699.LI Bing, LIU Hong, LI Youming.3-D seismic data discrete smooth interpolation using conjugate gradient[J].Chinese Journal of Geophysics, 2002, 45(5):691-699.
[14]
郭志宏.双三次样条内插网格化方法软件的研制开发及其应用[M].北京:地质出版社,1999.
[15]
刘沈衡, 李爱勇.分形分析在重磁不规则网网格化中的应用[J].石油物探, 1999, 38(3):108-111.LIU Shenheng, LI Aiyong.Application of fractal ana-lysis to gridding of irregular-grid gravity and magnetic data[J].Geophysical Prospecting for Petroleum, 1999, 38(3):108-111.
[16]
闫浩飞,刘国峰,薛典军,等.基于凸集投影方法的重磁数据规则缺失重建[J].地球物理学进展,2016,31(5):2192-2197.YAN Haofei, LIU Guofeng, XUE Dianjun, et al.Reconstruction of gravity/magnetic data with the projection-onto-convex-sets methods[J].Progress in Geophysics, 2016, 31(5):2192-2197.
[17]
SAMBRIDGE M, Braun J, Mcqueen H.Geophysical parametrization and interpolation of irregular data using natural neighbours[J].Geophysical Journal International, 2010, 122(3):837-857.
[18]
穆石敏.区域地球物理数据处理方法及其应用[M].吉林长春:吉林科学出版社,1980.
[19]
董焕成,李宗杰.重力异常匹配滤波的原理与应用[J].物探与化探,1993,17(5):331-340.DONG Huancheng, LI Zongjie.The principle and application of the matchfiltering method for gravitational anomalies[J].Geophysical and Geochemical Exploration, 1993, 17(5):331-340.
[20]
杨文采, 施志群, 侯遵泽, 等.离散小波变换与重力异常多重分解[J].地球物理学报, 2001,44(4):534-541,582.YANG Wencai,SHI Zhiqun,HOU Zunze, et al.Discrete wavelet transform for multiple decomposition of gravity anomalies[J].Chinese Journal of Geophysics, 2001, 44(4):534-541, 582.
[21]
郭志宏, 徐昆.中国航磁异常管理信息系统开发与建设[J].地球物理学进展, 2002, 17(4):703-710.GUO Zhihong, XU Kun.The development and construction of China aeromagnetic anomaly management information system of China[J].Progress in Geophy-sics, 2002, 17(4):703-710.
[22]
FERNANDO J S, SILVA D, VALÉRIA C F, et al.Adaptive learning 3D gravity inversion for salt-body imaging[J].Geophysics, 2011, 76(3):A49-A52.
[23]
韩卫雪, 周亚同, 池越.基于深度学习卷积神经网络的地震数据随机噪声去除[J].石油物探,2018,57(6):862-869, 877.HAN Weixue, ZHOU Yatong, CHI Yue.Deep learning convolutional neural networks for random noise attenuation in seismic data[J].Geophysical Prospecting for Petroleum, 2018, 57(6):862-869, 877.
[24]
王文强, 孟凡顺, 孙文亮.优化卷积神经网络在道编辑中的应用[J].地球物理学展, 2019, 34(1):214-220.WANG Wenqiang, MENG Fanshun, SUN Wenliang.Trace editing based on optimized convolutional neural network[J].Progress in Geophysics, 2019, 34(1):214-220.
[25]
王静.基于生成对抗网络的地震数据插值[D].黑龙江哈尔滨:哈尔滨工业大学, 2019.
[26]
WANG F, CHEN S C, and LIU Y N.Deep learning for gravity and magnetic data interpolation[C].SEG Technical Program Expanded Abstracts, 2019, 38:2533-2537.
[27]
郑浩, 张兵.基于卷积神经网络的智能化地震数据插值技术[J].地球物理学进展, 2020, 35(2):721-727.ZHENG Hao, ZHANG Bing.Intelligent seismic data interpolation via convolutional neural network[J].Progress in Geophysics, 2020, 35(2):721-727.
[28]
毛博.基于卷积神经网络的地震数据重构与模型构建研究[D].吉林长春:吉林大学, 2020.
[29]
陈贝.基于深度卷积神经网络的位场数据去噪研究[D].北京:中国地质大学(北京), 2020.
[30]
闫星宇, 顾汉明, 罗红梅, 等.基于改进深度学习方法的地震相智能识别[J].石油地球物理勘探,2020, 55(6):1169-1177.YAN Xingyu, GU Hanming, LUO Hongmei, et al.Intelligent seismic facies classification based on an improved deep learning method[J].Oil Geophysical Prospecting, 2020, 55(6):1169-1177.
[31]
李海山, 陈德武, 吴杰, 等.叠前随机噪声深度残差网络压制方法[J].石油地球物理勘探,2020,55(3):493-503.LI Haishan, CHEN Dewu, and WU Jie, et al.Pre-stack random noise suppression with deep residual network[J].Oil Geophysical Prospecting, 2020, 55(3):493-503.
[32]
伊小蝶, 吴帮玉, 孟德林, 等.数据增广和主动学习在波阻抗反演中的应用[J].石油地球物理勘探,2021, 56(4):707-715.YI Xiaodie, WU Bangyu, MENG Delin, et al.Application of data augumentation and active learning to seismic wave impendence inversion[J].Oil Geophysical Prospecting, 2021,56(4):707-715.
[33]
张岩,李新月,王斌,等.基于联合深度学习的地震数据随机噪声压制[J].石油地球物理勘探,2021,56(1):9-25.ZHANG Yan, LI Xinyue, WANG Bin,et al.Random noise suppression of seismic based on joint deep lear-ning[J].Oil Geophysical Prospecting, 2021, 56(1):9-25.
[34]
安鹏, 曹丹平.基于深度学习的测井岩性识别方法研究与应用[J].地球物理学进展, 2018, 33(3):1029-1034.AN Peng, CAO Danping.Research and application of logging lithology identification based on deep learning[J].Progress in Geophysics, 2018, 33(3):1029-1034.
[35]
ROBERTO F, RICCARDO B, RICHARD L.Mode-ling irregular small bodies gravity field via extreme learning machines and Bayesian optimization[J].Advances in Space Research, 2021, 67(1):617-638.
[36]
MNIH V, HEESS N, GRAVES A, et al.Recurrent models of visual attention[J].Advances in Neural Information Processing Systems, 2014, 3(1):121-128.
[37]
BAHDANAU D, CHO K, BENGIO Y.Neural Machine Translated by Jointly Learning to Align and Translate[OL].http://pdfs.semanticscholar.org/071b/16f25117fb6133480c6259227d54fc2a5ea0.pdf.
[38]
LUONG M T, PHAM H,MACHING C D.Effective Approaches to Attention Based Neural Machine Translation[OL].http://aclweb.org/anthology/D15-1166.
[39]
VASWANI A, SHAZEER N, PARMAR N, et al.Attention Is All You Need[M].arXiv, 2017.
