Deep learning-driven velocity modeling based on seismic reflection data and multi-scale training sets
HAN Mingliang1, ZOU Zhihui1,2, MA Rui1,3
1. Key Lab of Submarine Geosciences and Prospecting Techniques, MOE, College of Marine Geosciences, Ocean University of China, Qingdao, Shandong 266100, China; 2. Evaluation and Detection Technology Laboratory of Marine Mineral Resources, Qingdao National Laboratory for Marine Science and Technology, Qingdao, Shandong 266061, China; 3. Shanghai Transwarp Co. Ltd, Shanghai 200030, China
Abstract:With the enlargement of seismic data volume, conventional methods of seismic velocity modeling are facing challenges to stability, accuracy, and computational efficiency. A deep learning-driven velocity modeling method based on seismic reflection data and multi-scale training sets was proposed in this paper. This method combined reflection waveforms and velocity spectra into the input of a full convolutional neural network and adopted the dropout layer in the neural network to improve the generalization ability. Moreover, it integrated multi-scale training sets to realize the mapping from seismic data to velocity models. To test the effectiveness and applicability of this method in different geologic structures, numerical experiments were carried out with layered models, isolated anomaly models, and the BP salt dome models. Experimental results show when seismic reflection waveforms and velocity spectra are combined as the feature dataset of deep learning, the accuracy of velocity modeling is higher than that in the case where they were adopted individually. It overcomes the defects of unstable modeling caused by using reflection waveform alone and low modeling accuracy induced by using velocity spectra alone. Furthermore, the accuracy of velocity modeling results in anomaly boundaries with the multi-scale velocity model to construct the training set is higher than that with a single-scale model as the training set. After only one training process, the deep neural network can quickly build the velocity model of the underground structure which is similar to the velocity structure in the training set. Therefore, compared with the conventional method, it has higher computational efficiency and deserves promotion when building large amounts of velocity models.
韩明亮, 邹志辉, 马锐. 利用反射地震资料和多尺度训练集的深度学习速度建模[J]. 石油地球物理勘探, 2021, 56(5): 935-946.
HAN Mingliang, ZOU Zhihui, MA Rui. Deep learning-driven velocity modeling based on seismic reflection data and multi-scale training sets. Oil Geophysical Prospecting, 2021, 56(5): 935-946.
潘宏勋, 方伍宝.地震速度分析方法综述[J]. 勘探地球物理进展, 2006, 29(5):305-311.PAN Hongxun, FANG Wubao.Review of seismic velocity analysis methods[J]. Progress in Exploration Geophysics, 2006, 29(5):305-311.
[2]
张敏, 李振春.偏移速度分析与建模方法综述[J]. 勘探地球物理进展, 2007, 30(6):421-427.ZHANG Min, LI Zhenchun.Review of migration velocity model building[J]. Progress in Exploration Geo-physics, 2007, 30(6):421-427.
[3]
雷栋, 胡祥云.地震层析成像方法综述[J]. 地震研究, 2006, 29(4):418-426.LEI Dong, HU Xiangyun.Review of seismic tomography methods[J]. Journal of Seismological Research, 2006, 29(4):418-426.
[4]
李慧, 成德安, 金婧.网格层析成像速度建模方法与应用[J]. 石油地球物理勘探, 2013, 48(增刊1):12-16.LI Hui, CHENG Dean, JIN Jing.Velocity model building based on grid tomography[J]. Oil Geophy-sical Prospecting, 2013, 48(S1):12-16.
[5]
马彦彦, 李国发, 张星宇, 等.叠前深度偏移速度建模方法分析[J]. 石油地球物理勘探, 2014, 49(4):687-693.MA Yanyan, LI Guofa, ZHANG Xingyu, et al.Strategy of velocity model building in prestack depth migration[J]. Oil Geophysical Prospecting, 2014, 49(4):687-693.
[6]
韩明亮, 孙文亮, 邹志辉, 等.基于地震节点和测线数据联合的含油气盆地层析速度建模[C]. 中国石油学会2019年物探技术研讨会论文集, 2019, 1351-1354.HAN Mingliang, SUN Wenliang, ZOU Zhihui, et al.Velocity model building of oil-gas bearing basin based on joint data of seismic nodes and lines[C]. CPS 2019 Geophysical Exploration Technology Conference, 2019, 1351-1354.
[7]
Sun W L, Zou Z H, Zhou H W, et al.Velocity model building based on multiscale deformable-layer tomography using first-arrival and reflection traveltimes[C]. SEG Technical Program Expanded Abstracts, 2018, 37:5228-5232.
[8]
Wremez L, Lau K W H, Nedimovic M R, et al.Tra-veltime tomography of a dense wide-angle profile across Orphan Basin[J]. Geophysics, 2015, 80(3):B69-B82.
[9]
Zou Z H, Zhou H W, Gurrola H, et al.Impact and solutions of seawater heterogeneity on wide-angle tomographic inversion of crustal velocities in deep marine environments-numerical studies[J]. Journal of Earth Science, 2018, 29(6):1380-1389.
[10]
Tarantola A.Inversion of seismic reflection data in the acoustic approximation[J]. Geophysics, 1984, 49(8):1259-1266.
[11]
杨勤勇, 胡光辉, 王立歆.全波形反演研究现状及发展趋势[J]. 石油物探, 2014, 53(1):77-83.YANG Qinyong, HU Guanghui, WANG Lixin.Research status and development trend of full waveform inversion[J]. Geophysical Prospecting for Petroleum, 2014, 53(1):77-83.
[12]
王庆, 张建中, 黄忠来.时间域地震全波形反演方法进展[J]. 地球物理学进展, 2015, 30(6):2797-2806.WANG Qing, ZHANG Jianzhong, Huang Zhonglai.Progress in the time-domain full waveform inversion[J]. Progress in Geophysics, 2015, 30(6):2797-2806.
[13]
卞爱飞, 於文辉, 周华伟.频率域全波形反演方法研究进展[J]. 地球物理学进展, 2010, 25(3):982-993.BIAN Aifei, YU Wenhui, ZHOU Huawei.Progress in the frequency-domain full waveform inversion method[J]. Progress in Geophysics, 2010, 25(3):982-993.
[14]
杨午阳, 王西文, 雍学善, 等.地震全波形反演方法研究综述[J]. 地球物理学进展, 2013, 28(2):766-776.YANG Wuyang, WANG Xiwen, YONG Xueshan, et al.The review of seismic full waveform inversion method[J]. Progress in Geophysics, 2013, 28(2):766-776.
[15]
Araya-polo M, Jennings J, Adler A, et al.Deep-lear-ning tomography[J]. The Leading Edge, 2018, 37(1):58-66.
[16]
Mosser L, Kimman W, Dramsch J, et al.Rapid seismic domain transfer:Seismic velocity inversion and mode-ling using deep generative neural networks[C]. Extended Abstracts of 80th EAGE Conference & Exhibition, 2018, PP1.
[17]
Li S, Liu B, Ren Y, et al.Deep learning inversion of seismic data[J]. IEEE Transactions on Image Processing, 2019, 58(3):2135-2149.
[18]
Zhang Z P, Yue W, Zheng Z, et al.VelocityGAN:subsurface velocity image estimation using conditio-nal adversarial networks[C]. IEEE Winter Conference on Applications of Computer Vision, 2019, 705-714.
[19]
Yuan C C, Zhang X, Jia X F, et al.Time-lapse velocity imaging via deep learning[J]. Geophysical Journal International, 2020, 220(2):1228-1241.
[20]
Yang F S, Ma J W.Deep-learning inversion:a next generation seismic velocity-model building method[J]. Geophysics, 2019, 84(4):R583-R599.
[21]
Park M J, Sacchi M D.Automatic velocity analysis using convolutional neural network and transfer lear-ning[J]. Geophysics, 2020, 85(1):V33-V43.
Wang W L, Ma J W.Velocity model building in a crosswell acquisition geometry with image-trained arti-ficial neural networks[J]. Geophysics, 2020, 85(2):U31-U46.
[24]
Chen Y Q, Schuster G T.Seismic inversion by newtonian machine learning[J]. Geophysics, 2020, 85(4):WA185-WA200.
[25]
孙宇航, 刘洋.利用GRU神经网络预测横波速度[J]. 石油地球物理勘探, 2020, 55(3):484-492, 503.SUN Yuhang, LIU Yang.Prediction of S-wave velocity based on GRU neural network[J]. Oil Geophysical Prospecting, 2020, 55(3):484-492, 503.
[26]
张昊, 朱培民, 顾元, 等.基于深度学习的地震速度谱自动拾取方法[J]. 石油物探, 2019, 58(5):724-733.ZHANG Hao, ZHU Peimin, GU Yuan, et al.Velocity auto-picking from velocity spectra based on deep learning[J]. Geophysical Prospecting for Petroleum, 2019, 58(5):724-733.
[27]
张玉玺, 刘洋, 张浩然, 等.基于深度学习的多属性盐丘自动识别方法[J]. 石油地球物理勘探, 2020, 55(3):475-483.ZHANG Yuxi, LIU Yang, ZHANG Haoran, et al.Multi-attribute automatic interpretation of salt domes based on deep learning[J]. Oil Geophysical Prospecting, 2020, 55(3):475-483.
[28]
Zhou H W.Multiscale traveltime tomography[J]. Geo-physics, 2003, 68(5):1639-1649.
[29]
Long J, Shelhamer E, Darrel T.Fully convolutional networks for semantic segmentation[C]. Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 2015, 3431-3440.
[30]
Ronneberger O, Fischer P, Brox T.U-Net:Convolutional networks for biomedical image segmentation[C]. International Conference on Medical Image Computing and Computer-Assisted Intervention, 2015, 234-241.
[31]
高雪鹏, 丛爽.BP网络改进算法的性能对比研究[J]. 控制与决策, 2001, 16(2):167-171.GAO Xuepeng, CONG Shuang.Comparative study on fast learning algorithms of BP networks[J]. Control and Decision, 2001, 16(2):167-171.
[32]
Billette F J, Brandsberg-dhal S.The 2004 BP velocity benchmark[C]. Extended Abstracts of 67th EAGE Conference & Exhibition, 2005, B035.
[33]
Kosloff D D, Baysal E.Forward modeling by a Fourier method[J]. Geophysics, 1982, 47(10):1402-1412.
[34]
马锐, 邹志辉, 芮拥军, 等.基于SPML和海绵边界的伪谱法弹性波模拟复合吸收边界条件[J]. 石油物探, 2018, 57(1):94-103.MA Rui, ZOU Zhihui, RUI Yongjun, et al.A compo-site absorbing boundary based on the SPML and sponge absorbing boundary for pseudo-spectral elastic wave modeling[J]. Geophysical Prospecting for Petroleum, 2018, 57(1):94-103.