基于混合优化与改进的U-Net震源分离方法

A Seismic Source Separation Method Based on Hybrid Optimization and Improved U-Net

摘要

传统单震源地震勘探存在效率低下和抗干扰能力不足的问题，而多震源技术虽然提高了勘探效率，但因混叠噪声的干扰导致数据质量下降。为此，提出了两种优化方法以解决震源分离问题。方法一：通过融合FISTA算法与ALBM算法构建动态加权混合优化算法（ALFT），在保证精度的同时提升了收敛速度，并结合滤波法与反演法的优势，形成了“初值预判-迭代修正”的流程。实验结果表明，相较于直接迭代方法，该方法可使信噪比提升10%~25%，迭代时间减少33%。方法二：提出了一种CSA-Unet深度学习网络模型，该模型基于U-Net网络架构，引入注意力局部对比度模块以增强对有效信号特征的捕获能力，并结合局部熵离散点抑制机制剔除辅震源干扰。验证结果显示，无论是在模拟数据集（Sigsbee2B）还是真实数据集上，CSA-UNet的分离信噪比明显优于ALFT_a和U-Net，同时有效保护了地层反射信号结构。本文所提出的方法为多震源地震勘探提供了高效且高精度的解决方案，在复杂地质条件下的成像应用中具有重要意义。

Abstract

Traditional single-source seismic exploration has problems of low efficiency and insufficient anti-interference ability. Although multi-source technology improves the exploration efficiency, the data quality deteriorates due to the interference of aliasing noise. For this reason, this paper proposes two optimization methods to solve the source separation problem. Method 1: A dynamic weighted hybrid optimization algorithm (ALFT) is constructed by integrating the FISTA algorithm and the ALBM algorithm. This algorithm improves the convergence speed while ensuring accuracy. By combining the advantages of the filtering method and the inversion method, a process of “initial value pre-judgment-iterative correction” is formed. The experimental results show that, compared with the direct iteration method, this method can increase the signal-to-noise ratio by 10%~25% and reduce the iteration time by 33%. Method 2: A CSA–Unet deep learning network model is proposed. Based on the U–Net network architecture, this model introduces an attention local contrast (ALC) module to enhance the ability to capture the characteristics of effective signals, and combines a local entropy discrete point suppression mechanism to eliminate the interference of auxiliary sources. The validation results demonstrate that CSA–UNet achieves a significantly higher separation signal-to-noise ratio than ALFT_a and U–Net on both the simulated dataset (Sigsbee2B) and the real dataset, while also effectively preserving the structure of the formation reflection signals. The methods proposed in this paper provide an efficient and high-precision solution for multi-source seismic exploration and are of great significance in imaging practices under complex geological conditions.

关键词

多震源地震勘探 / 震源混叠噪声 / 主辅震源分离 / U-Net

Key words

multi-source seismic exploration / source aliasing noise / separation of primary and auxiliary sources / U–Net

参考文献

[1] 庞雄奇. 油气田勘探[M]. 北京：石油工业出版社， 2006.

[2] PANG Xiongqi. Oil and gas field exploration[M]. Beijing: Petroleum Industry Press, 2006.

[3] YILMAZ Ö . Seismic data analysis: Processing, inversion, and interpretation of seismic data[M]. Tulsa: Society of Exploration Geophysicists, 2001.

[4] BEASLEY C J, CHAMBERS R E, JIANG Z R . A new look at simultaneous sources[C]. SEG Technical Program Expanded Abstracts, 1998. doi: 10.1190/1.1820149

[5] SILVERMAN D. Method of three-dimensional seismic prospecting: US19770844355[P]. 1979-06-26.

[6] MUELLER M B, HALLIDAY D F, MANEN D V, et al. The benefit of encoded source sequences for simultaneous source separation[J]. Geophysics, 2015, 80(5): 133-143. doi: 10.1190/geo2015-0015.1

[7] ZHAO Hu, WU Sihai, YANG Jing, et al. Designing optimal number of receiving traces based on simulation model[J]. Applied Geophysics, 2017, 14(1): 49-55. doi: 10.1007/s11770-017-0603-3

[8] CHEN Yangkang, CHEN Hanming, XIANG Kui, et al. Preserving the discontinuities in least-squares reverse time migration of simultaneous-source data[J]. Geophysics, 2017, 82(3): 185-196. doi: 10.1190/geo2016-0456.1

[9] BAI Min, WU Juan, XIE Jianyong, et al. Least-squares reverse time migration of blended data with low-rank constraint along structural direction[J]. Journal of Seismic Exploration, 2018, 27(1): 29-48.

[10] HUANG Weilin, WANG Runqiu, GONG Xiangbo, et al. Iterative deblending of simultaneous-source seismic data with structuring median constraint[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(1): 58-62. doi: 10.1109/LGRS.2017.2772857

[11] LIU Wei, CHEN Wei . Recent advancements in empirical wavelet transform and its applications[J]. IEEE Access, 2019, 7: 103770-103780. doi: 10.1109/ACCESS.2019.2930529

[12] CHEN Yangkang . Deblending using a space-varying median filter[J]. Exploration Geophysics, 2015, 46(4): 332-341. doi: 10.1071/EG14051

[13] GAN Shuwei, WANG Shoudong, CHEN Yangkang, et al. Simultaneous-source separation using iterative seislet-frame thresholding[J]. IEEE Geoscience and Remote Sensing Letters, 2016, 13(2): 197-201. doi: 10.1109/LGRS.2015.2505319

[14] ZU Shaohuan, ZHOU Hui, CHEN Yangkang, et al. A marine field trial for iterative deblending of simultaneous sources[C]. Dallas: SSEG International Exposition and 86th Annual Meeting, 2016.

[15] ZU Shaohuan, ZHOU Hui, MAO Weijian, et al. Iterative deblending of simultaneous-source data using a coherency-pass shaping operator[J]. Geophysical Journal International, 2017, 211(1): 541-557. doi: 10.1093/GJI/GGX324

[16] VAN B R, BAARDMAN R, MARTIN T, et al. An inversion approach to separating sources in marine simultaneous shooting acquisition-application to a Gulf of Mexico data set[J]. Geophysical Prospecting, 2012, 60(4): 640-647. doi: 10.1111/j.1365-2478.2012.01076.x

[17] ZU Shaohuan, ZHOU Hui, CHEN Yangkang, et al. A periodically varying code for improving deblending of simultaneous sources in marine acquisition[J]. Geophysics, 2016, 81(3): 213-225. doi: 10.1190/GEO2015-0447.1

[18] QU Shan, ZHOU Hui, LIU Renwu, et al. Deblending of simultaneous-source seismic data using fast iterative shrinkage-thresholding algorithm with firm thresholding[J]. Acta Geophysica, 2016, 64(4): 1064-1092. doi: 10.1515/acgeo-2016-0043

[19] ZHANG Hua, LIANG Shuang, PENG Qing, et al. Separation of simultaneous source data via an accelerated joint iterative method[J]. Journal of Applied Geophysics, 2022, 203: 104680. doi: 10.1016/j.jappgeo.2022.104680

[20] SHAO T, HU W, PHAN S, et al. Time-lapse seismic data deblending with deep learning[C]. Vienna: 84th EAGE Annual Conference & Exhibition, 2023. doi: 10.3997/2214-4609.202310334

[21] MU Shengqiang, LI Wenda, WU Tianqi, et al. A mixed domain deblending approach for simultaneous source data based on deep learning[J]. Journal of Applied Geophysics, 2024, 228: 105451. doi: 10.1016/j.jappgeo.2024.106785

[22] MA Liyun, HAN Liguo, ZHANG Pan . Iterative separation of blended seismic data in shot domain using deep learning[J]. Remote Sensing, 2024, 16(22): 4167. doi: 10.3390/rs16224167

[23] LI Ji, TRAD D, LIU Dawei. Robust seismic data denoising via self-supervised deep learning[J]. Geophysics, 2024, 89(5): 437-451. doi: 10.1190/geo2023-0762.1

[24] DUAN Zhongyu, WANG Haojie, WANG Yunlei . Simultaneous source separation and noise reduction with structure-oriented space-varying alpha-trimmed mean filter[J]. Journal of Applied Geophysics, 2024, 220: 105253. doi: 10.1016/j.jappgeo.2023.105253

[25] 杨熙熙， 张华， 武召祺， 等. 基于曲波变换与快速指数阈值迭代的多震源地震数据分离[J]. 科学技术与工程， 2024， 24（13）： 5296-5303. doi: 10.12404/j.issn.1671-1815.2306283

[26] YANG Xixi, ZHANG Hua, WU Zhaoqi, et al. Separation of multi-source seismic data based on curvelet-transform and fast exponential threshold iteration method[J]. Science Technology and Engineering, 2024, 24(13): 5296-5303. doi: 10.12404/j.issn.1671-1815.2306283

[27] HUANG B, MA S, GOLDFARB D . Accelerated linearized bregman method[J]. Journal of Scientific Computing, 2013, 54(2): 428-453. doi: 10.1007/s10915-012-9592-9

基金资助

中国石油西南石油大学创新联合体支持交叉学科发展“揭榜挂帅”项目(2024CXJB09)

新型油气勘探开发国家科技重大专项(2025ZD1408800)