Comparative analysis of three seismic impedance inversion methods based on deep learning
WANG Zefeng1, LI Yonggen2, XU Huiqun1, YANG Mengqiong1, ZHAO Yasong1, PENG Zhen1
1. College of Geophysics and Petroleum Resources, Yangtze University, Wuhan, Hubei 430100, China; 2. Research Institute of Petroleum Exploration and Development, PetroChina, Beijing 100083, China
Abstract：The difference in neural network structure leads to different deep learning effects. Hence, upon the comparison of the fully convolutional neural network (FCN), convolutional recurrent neural network (CRNN), and time-domain convolutional neural network (TCN), this study uses the forward model tests to comparatively analyze the accuracy and computational efficiency of seismic impedance inversion methods based on the above three deep learning methods. Moreover, the three methods are applied to actual data for further comparison. The experimental results show that the computational efficiency and accuracy of TCN-based wave impedance inversion are relatively high. For wave impedance inversion based on TCN, FCN, and CRNN, the inversion time is 82 s, 68 s, and 264 s, respectively, and the inversion accuracy is 99.15%, 97.84%, and 98.14%, respectively. The actual data application reveals that the results of TCN-based wave impedance inversion match better with the logging data. This conclusion can provide a reference for the optimization and selection of intelligent wave impedance inversion methods.
LI M,LI Y,WU N,et al. Desert seismic random noise reduction framework based on improved PSO-SVM[J]. Acta Geodaetica et Geophysica,2020,55(1): 101-117.
[2]
谭锋奇,李洪奇,许长福,等. 基于聚类分析方法的砾岩油藏储层类型划分[J]. 地球物理学进展,2012,27(1): 246-254.TAN Fengqi,LI Hongqi,XU Changfu,et al. Reservoir classification of conglomerate reservoir base on clustering analysis method[J]. Progress in Geophy-sics,2012,27(1): 246-254.
[3]
顾元,朱培民,荣辉,等. 基于贝叶斯网络的地震相分类[J]. 地球科学(中国地质大学学报),2013,38(5): 1143-1152.GU Yuan,ZHU Peimin,RONG Hui,et al. Seismic facies classification based on Bayesian networks[J]. Earth Science (Journal of China University of Geosciences),2013,38(5): 1143-1152.
[4]
李晓光,吴潇. 从SEG年会看人工智能在地震数据处理与解释中的新进展[J]. 世界石油工业,2020,27(4): 27-35.LI Xiaoguang,WU Xiao. Progresses of artificial intelligence on seismic data processing and interpretation reviewed from SEG annual meetings[J]. World Petroleum Industry,2020,27(4): 27-35.
王延光,李皓,李国发,等. 一种用于薄层和薄互层砂体厚度估算的复合地震属性[J]. 石油地球物理勘探,2020,55(1): 153-160.WANG Yanguang,LI Hao,LI Guofa,et al. A composite seismic attribute used to estimate the sand thickness for thin bed and thin interbed[J]. Oil Geophysical Prospecting,2020,55(1): 153-160.
[7]
刘艳,赵海涛,徐红霞,等. 利用叠后提频及多属性分析预测砂泥岩薄互层——以轮南石炭系砂泥岩段为例[J]. 石油地球物理勘探,2018,53(增刊1): 196-200.LIU Yan,ZHAO Haitao,XU Hongxia,et al. Thin inter-bed prediction with poststack frequency improvement and multi-attribute analysis: an example of Carboniferous sand-mudstone member in Lunnan area[J]. Oil Geophysical Prospecting,2018,53(S1): 196-200.
[8]
BOADU F K. Inversion of fracture density from field seismic velocities using artificial neural networks[J]. Geophysics,1998,63(2): 534-545.
[9]
李薇薇,龚仁彬,周相广,等. 基于深度学习UNet++网络的初至波拾取方法[J]. 地球物理学进展,2021,36(1): 187-194.LI Weiwei,GONG Renbin,ZHOU Xiangguang,et al. UNet++: a deep-neural-network-based seismic arrival time picking method[J]. Progress in Geophysics,2021,36(1): 187-194.
[10]
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.
[11]
马江涛,刘洋,张浩然. 地震相智能识别研究进展[J]. 石油物探,2022,61(2): 262-275.MA Jiangtao,LIU Yang,ZHANG Haoran. Research progress on intelligent identification of seismic facies[J]. Geophysical Prospecting for Petroleum,2022,61(2): 262-275.
[12]
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.
李庆忠. 论地震约束反演的策略[J]. 石油地球物理勘探,1998,33(4): 423-438.LI Qingzhong. On strategy of seismic restricted inversion[J]. Oil Geophysical Prospecting,1998,33(4): 423-438.
[16]
LECUN Y,BOTTOU L,BENGIO Y,et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE,1998,86(11): 2278-2324.
[17]
ZAREMBA W,SUTSKEVER I,VINYALS O. Recurrent neural network regularization[DB/OL]. (2015-02-19)[2021-10-04]. https://arxiv.org/abs/1409.2329.
[18]
DAS V,POLLACK A,WOLLNER U,et al. Convolutional neural network for seismic impedance inve-rsion[J]. Geophysics,2019,84(6): R869-R880.
[19]
WU B Y,MENG D L,WANG L L,et al. Seismic impedance inversion using fully convolutional residual network and transfer learning[J]. IEEE Geoscience and Remote Sensing Letters,2020,17(12): 2140-2144.
[20]
ALFARRAJ M,ALREGIB G. Semi-supervised lear-ning for acoustic impedance inversion[C]. SEG Technical Program Expanded Abstracts,2019,38: 2298-2302.
[21]
WU B Y,MENG D L,ZHAO H X. Semi-supervised learning for seismic impedance inversion using generative adversarial networks[J]. Remote Sensing,2021,13(5): 909.
[22]
MUSTAFA A,ALFARRAJ M,ALREGIB G. Estimation of acoustic impedance from seismic data using temporal convolutional network[C]. SEG Technical Program Expanded Abstracts,2019,38: 2554-2558.
[23]
SHELHAMER E,LONG J,DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(4): 640-651.
[24]
SHI B G,BAI X,YAO C. An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2017,39(11): 2298-2304.
[25]
BAI S J,KOLTER J Z,KOLTUN V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[DB/OL]. (2018-04-19)[2021-10-04]. https://arxiv.org/abs/1803.01271.
