王尚红,彭海源,祝新玉,李军权,张宝林

• 1:国家能源集团新疆能源化工有限公司昌吉矿业分公司
• 2:中煤科工西安研究院(集团)有限公司

摘要(Abstract)：

针对深度学习去噪过程中易导致槽波色散特征畸变、Airy相群时延偏移及多道传播不连续等问题，构建了一种以物理属性恢复为导向的煤层槽波信号增强方法。首先，基于分层介质导波理论，确定频率-波数域与频率-时延域的理论约束边界；随后将色散一致性、群时延偏差、跨道相干性及残余能量作为正则项写入目标函数，实现物理规律对卷积自编码网络训练的显式引导；最后通过无噪参考、常规噪声、强非平稳干扰、参数失配、消融试验及类现场盲测等多组试验，系统检验方法对槽波物理量恢复及后续解释任务的影响。结果表明：所提方法在低信噪比环境下能够稳定恢复槽波主模态，保持Airy相主频带的动力学特征，增强多道传播连续性，并使输出能量在频率-波数域内更集中于理论允许区。与带通滤波、F-K滤波及纯数据驱动自编码器相比，该方法在波形保持、Airy相时延控制、物理域外能量抑制以及断层定位与层析输入精度方面均表现出更优结果。

关键词(KeyWords)： 煤层槽波;物理信息深度学习;色散保持去噪;Airy相;断层定位;层析输入

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金面上项目（42074175）;; 陕西省自然科学基金项目（2022JM-159）

作者(Author): 王尚红,彭海源,祝新玉,李军权,张宝林

参考文献(References)：

• [1] ZHANG J, YANG C, LIU S, et al. Detection of geological anomalies in coal mining working faces using a scattered-wave imaging method[J]. Journal of Petroleum Exploration and Production Technology, 2023(13):1299-1313.
• [2] CZARNY R, MALINOWSKI M, CHAMARCZUK M, et al.Dispersive seismic waves in a coal seam around the roadway in the presence of excavation damaged zone[J]. International Journal of Rock Mechanics and Mining Sciences, 2021, 148.
• [3]姬广忠，刘克跃，李志勇，等.煤层工作面槽波3D频散反演成像研究[J].煤炭学报，2025, 50(12):5215-5222.JI G, LIU K, LI Z, et al. Study on 3D dispersion inversion imaging of channel waves in coal seam working faces[J]. Journal of China Coal Society, 2025, 50(12):5215-5222.
• [4] WANG W, GAO X, WU Y. A Study on the Imaging Method for the Channel Wave Dispersion curve Variability Function[J]. Minerals,2023, 13(1).
• [5] SONG J, ZHOU S, LIU Q, et al. Guided wave signal-based sensing and classification for small geological structure detection in coal mines[J]. IET Signal Processing, 2023, 17(7).
• [6]王勃，申思宏任，蔚立元，等.基于VMD和GA-SVM的矿井地震自适应噪声压制方法[J].煤炭学报，2024, 49(3):1530-1538.WANG B, SHENSI H, YU L, et al. Adaptive noise suppression method for mine seismic signals based on VMD and GA-SVM[J]. Journal of China Coal Society, 2024, 49(3):1530-1538.
• [7] QIN K, et al. High-precision coal mine microseismic P-wave arrival picking via physics-constrained deep learning[J]. Sensors,2025, 25(23).
• [8]周官群，薛凯文，张维鑫，等.基于CNN-LSTM的透射槽波勘探走向断层预测研究[J].工矿自动化，2025, 51(7).ZHOU G, XUE K, ZHANG W, et al. Research on strike fault prediction of transmission channel wave exploration based on CNNLSTM[J]. Industry and Mine Automation, 2025, 51(7).
• [9] DING M, et al. Seismic signal denoising using Swin-Conv-UNet[J]. Journal of Applied Geophysics, 2024, 225.
• [10] ZHONG T, LI X, ZHANG J, et al. Seismic random noise suppression by using deep residual U-Net[J]. Journal of Petroleum Science and Engineering, 2022, 210.
• [11] ZHANG Y, et al. Seismic data denoising with two-step prediction strategy based on improved Dn CNN and U-Net[J]. Computers&Geosciences, 2024, 185.
• [12] WANG F, et al. Seismic Denoising by deep learning from natural repeating earthquakes[J]. Journal of Geophysical Research:Solid Earth, 2025, 130(1).
• [13] MANDELLI S, et al. A convolutional autoencoder method for simultaneous seismic data reconstruction and denoising[J].Geophysics, 2022, 87(1):43-54.
• [14]潘登，解志朝，杨志立，等.煤矿工作面背景噪声面波走时层析成像研究[J].煤炭学报，2025(12):5193-5204.PAN D, XIE Z, YANG Z, et al. Research on surface wave travel-time tomography of ambient noise in coal mine working faces[J]. Journal of China Coal Society, 2025(12):5193-5204.
• [15] LIU Y, et al. Seismic source location using deep learning and a metaheuristic-optimized grid search algorithm under a sparse network in underground coal mines[J]. Geomatics, Natural Hazards and Risk,2025, 16(1).
• [16] CHEN K, et al. Theoretical calculation of dispersion and attenuation curves of guided waves in layered media[J]. Geophysical Journal International, 2025, 240(2):393-410.