Prediction of strike-slip fault using transmitted in-seam wave exploration based on CNN-LSTM
-
摘要:
透射槽波地震勘探是探测工作面地质构造和灾害体的有效手段，但存在探测深度浅、分辨率低、易受地形与环境噪声干扰等问题。针对该问题，将深度学习技术引入透射槽波地震勘探，用于实现工作面走向断层位置预测。建立工作面走向断层地质模型，采用弹性波有限差分算法进行槽波正演模拟，生成槽波模拟数据集。构建卷积神经网络−长短期记忆（CNN−LSTM）网络模型，通过CNN提取槽波数据的局部特征，由LSTM网络捕捉槽波数据的时序依赖关系，实现槽波时空特征协同解析。采用槽波模拟数据集训练CNN−LSTM模型，预测的均方根误差为4.393 4 m，平均绝对误差为2.987 5 m，决定系数为0.988 3，验证了该模型具有较高的预测精度和较好的泛化能力。采用内蒙古某矿506工作面透射槽波勘探数据对CNN−LSTM模型进行迁移训练和验证，结果表明该模型预测的断层位置和走向与回采揭露的实际位置一致，预测效果优于槽波能量衰减成像、无线电坑透探测技术。
Abstract:Transmitted in-seam wave seismic exploration is an effective method for detecting geological structures and hazardous bodies in working faces, but it suffers from problems such as shallow exploration depth, low resolution, and susceptibility to terrain and environmental noise. To address these issues, deep learning technology was introduced into transmitted in-seam wave seismic exploration to predict the location of strike-slip faults in working faces. A geological model of the strike-slip fault in the working face was established, and the elastic wave finite difference algorithm was used to perform forward modeling of in-seam waves to generate a simulation dataset. A Convolutional Neural Network–Long Short-Term Memory (CNN-LSTM) model was constructed. CNN was used to extract local features of the in-seam wave data, and LSTM was used to capture the temporal dependencies in the in-seam wave data, thereby achieving collaborative analysis of spatiotemporal features. The CNN-LSTM model was trained using the in-seam wave simulation dataset. The predicted root mean square error was 4.393 4 m, the mean absolute error was 2.987 5 m, and the coefficient of determination was 0.988 3, verifying the model’s high prediction accuracy and good generalization capability. The CNN-LSTM model was then fine-tuned using transfer learning and validated using transmitted in-seam wave exploration data from the 506 working face of a mine in Inner Mongolia. The results showed that the predicted fault location and strike were consistent with the actual position revealed by mining, and the prediction performance was better than that of the in-seam wave energy attenuation imaging and radio borehole penetration detection technologies.
-
-
![]()
图 2 槽波激发点和接收点布置
Figure 2. Arrangement of excitation and receiving points of in-seam wave
![]()
图 4 基于CNN−LSTM模型的断层预测流程
Figure 4. Workflow of fault prediction based on CNN-LSTM model
![]()
图 10 506工作面槽波地震勘探现场布置
Figure 10. Site layout of in-seam wave seismic exploration in506 working face
![]()
图 11 槽波信号降噪前后波形对比
Figure 11. Comparison of waveform of in-seam wave data before and after noise reduction
![]()
图 13 第1种槽波勘探布置方式的断层构造预测结果
Figure 13. Tectonic fault prediction results for the 1st in-seam wave exploration arrangement
![]()
图 14 第1种槽波勘探布置方式的断层位置拟合结果
Figure 14. Fault location fitting results for the 1st in-seam wave exploration arrangement
![]()
图 15 第2种槽波勘探布置方式的断层构造预测结果
Figure 15. Tectonic fault prediction results for the 2nd in-seam wave exploration arrangement
![]()
图 16 第2种槽波勘探布置方式的断层位置拟合结果
Figure 16. Fault location fitting results for the 2nd in-seam wave exploration arrangement
表 1 煤层和围岩波速设置
Table 1 Wave velocity settings for coal seams and surrounding rock
m/s 编号 煤层波速 围岩波速 纵波 横波 纵波 横波 1 2 300 1 300 3 800 2 300 2 2 100 1 200 3 400 1 200 3 1 900 1 100 3 300 2 000 
下载: 导出CSV
表 2 实验平台硬件配置
Table 2 Hardware configurations of experimental platform
名称 型号 CPU Intel i7−9750H GPU NVIDIA 1660 Ti 内存 三星 8 GiB DDR4 2 667 MHz 磁盘 WDC PC SN720 （1 024 GiB）
下载: 导出CSV
表 3 模型评价结果
Table 3 Model evaluation results
数据集 RMSE/m MAE/m R2 训练集 4.393 4 2.987 5 0.988 3 测试集 5.732 2 3.840 6 0.986 5
下载: 导出CSV
-
[1] 胡泽安,张平松,许光泉. 煤层工作面地震波层析成像技术研究进展[J]. 地球物理学进展,2017,32(6):2451-2459. DOI: 10.6038/pg20170623 HU Zean,ZHANG Pingsong,XU Guangquan. Research advances of seismic tomography technology in coal seam[J]. Progress in Geophysics,2017,32(6):2451-2459. DOI: 10.6038/pg20170623
[2] 董守华,黄亚平,金学良,等. 煤田高密度三维地震勘探技术的发展现状及趋势[J]. 煤田地质与勘探,2023,51(2):273-282. DOI: 10.12363/issn.1001-1986.23.03.0116 DONG Shouhua,HUANG Yaping,JIN Xueliang,et al. Development status and trend of high-density 3D seismic exploration technology for coal fields[J]. Coal Geology & Exploration,2023,51(2):273-282. DOI: 10.12363/issn.1001-1986.23.03.0116
[3] 姬广忠,程建远,胡继武,等. 槽波衰减系数成像方法及其应用[J]. 煤炭学报,2014,39(增刊2):471-475. JI Guangzhong,CHENG Jianyuan,HU Jiwu,et al. Imaging method of slot wave attenuation coefficient and its application[J]. Journal of China Coal Society,2014,39(S2):471-475.
[4] 姬广忠,杨建华,张广学,等. 槽波振幅衰减系数CT成像技术应用[J]. 中国煤炭地质,2015,27(9):75-78. DOI: 10.3969/j.issn.1674-1803.2015.09.18 JI Guangzhong,YANG Jianhua,ZHANG Guangxue,et al. Application of in-seam wave amplitude attenuation coefficient CT imaging technology[J]. Coal Geology of China,2015,27(9):75-78. DOI: 10.3969/j.issn.1674-1803.2015.09.18
[5] 胡泽安,张平松,许光泉. 采煤工作面拟2.5维地震速度层析成像方法[J]. 煤炭学报,2018,43(9):2579-2586. HU Zean,ZHANG Pingsong,XU Guangquan. Pseudo 2.5 dimensional seismic velocity tomography in coal mining face[J]. Journal of China Coal Society,2018,43(9):2579-2586.
[6] 田瀚,吴荣新,胡泽安,等. 透射槽波相邻道质心频率的层析成像方法[J]. 煤田地质与勘探,2022,50(11):187-194. DOI: 10.12363/issn.1001-1986.22.04.0221 TIAN Han,WU Rongxin,HU Zean,et al. Tomography method for adjacent channel frequency shift of transmitted in-Seam waves[J]. Coal Geology & Exploration,2022,50(11):187-194. DOI: 10.12363/issn.1001-1986.22.04.0221
[7] HAN Jianguang,LYU Qingtian,ZHANG Zhiheng,et al. Application of Gaussian beam summation migration in reflected in-seam wave imaging[J]. Acta Geologica Sinica (English Edition),2024,98(1):276-284. DOI: 10.1111/1755-6724.15108
[8] WU Yanhui,ZHU Guowei,WANG Wei. Precise prediction of the collapse column based on channel wave spectral disparity characteristics and velocity tomography imaging[J]. Journal of Geophysics and Engineering,2022,19(3):326-335. DOI: 10.1093/jge/gxac019
[9] 郭银景,巨媛媛,范晓静,等. 槽波地震勘探研究进展[J]. 煤田地质与勘探,2020,48(2):216-227. DOI: 10.3969/j.issn.1001-1986.2020.02.032 GUO Yinjing,JU Yuanyuan,FAN Xiaojing,et al. Progress in research of in-seam seismic exploration[J]. Coal Geology & Exploration,2020,48(2):216-227. DOI: 10.3969/j.issn.1001-1986.2020.02.032
[10] 高墨通,杨维芳,刘祖昱,等. 结合卷积神经网络和注意力机制的LSTM采空区地表沉降预测方法[J]. 测绘通报,2024(6):53-58,170. GAO Motong,YANG Weifang,LIU Zuyu,et al. LSTM goaf surface subsidence prediction method combining convolutional neural network and attention mechanism[J]. Bulletin of Surveying and Mapping,2024(6):53-58,170.
[11] PENG Ping'an,HE Zhengxiang,WANG Liguan,et al. Automatic classification of microseismic records in underground mining:a deep learning approach[J]. IEEE Access,2020,8:17863-17876. DOI: 10.1109/ACCESS.2020.2967121
[12] LI Kewen,LI Xiao,YIN Ruonan,et al. A method for seismic fault identification based on self-training with high-stability pseudo-labels[J]. Applied Soft Computing,2024,163. DOI: 10.1016/J.ASOC.2024.111894.
[13] 姬广忠,吴荣新,张平松,等. 黏弹TI煤层介质3层模型Love槽波频散与衰减特征[J]. 煤炭学报,2021,46(2):566-577. JI Guangzhong,WU Rongxin,ZHANG Pingsong,et al. Dispersion and attenuation characteristics of Love channel waves in the three-layer model of viscoelastic TI coal seam media[J]. Journal of China Coal Society,2021,46(2):566-577.
[14] 姬广忠,魏久传,杨思通,等. HTI煤层介质槽波波场与频散特征初步研究[J]. 地球物理学报,2019,62(2):789-801. DOI: 10.6038/cjg2019M0230 JI Guangzhong,WEI Jiuchuan,YANG Sitong,et al. Preliminary study on wave field and dispersion characteristics of channel waves in HTI coal seam medium[J]. Chinese Journal of Geophysics,2019,62(2):789-801. DOI: 10.6038/cjg2019M0230
[15] 姬广忠,程建远,朱培民,等. 煤矿井下槽波三维数值模拟及频散分析[J]. 地球物理学报,2012,55(2):645-654. DOI: 10.6038/j.issn.0001-5733.2012.02.028 JI Guangzhong,CHENG Jianyuan,ZHU Peimin,et al. 3-D numerical simulation and dispersion analysis of in-seam wave in underground coal mine[J]. Chinese Journal of Geophysics,2012,55(2):645-654. DOI: 10.6038/j.issn.0001-5733.2012.02.028
[16] 王保利,金丹,张唤兰,等. 煤层中断层的透射槽波定量响应特征[J]. 煤炭学报,2022,47(8):2985-2991. WANG Baoli,JIN Dan,ZHANG Huanlan,et al. Quantitative response characteristics of transmitted in-seam channel wave in coal seam with faults[J]. Journal of China Coal Society,2022,47(8):2985-2991.
[17] 呼邦兵,朱国维,刘金锁,等. 掘进煤巷采空区Rayleigh型槽波超前探测三维数值模拟研究[J]. 煤炭工程,2020,52(2):121-125. HU Bangbing,ZHU Guowei,LIU Jinsuo,et al. Three-dimensional numerical simulation of the Rayleigh channel wave advance detection in the goaf of coal roadway[J]. Coal Engineering,2020,52(2):121-125.
[18] 韩建光,谷丙洛,王硕,等. 煤层反射槽波三维正演模拟[J]. 地质学报,2024,98(7):2258-2268. HAN Jianguang,GU Bingluo,WANG Shuo,et al. Study on 3D forward modeling of reflected in-seam wave in coal seam[J]. Acta Geologica Sinica,2024,98(7):2258-2268.
[19] LIU Bin,YANG Senlin,REN Yuxiao,et al. Deep-learning seismic full-waveform inversion for realistic structural models[J]. Geophysics,2021,86(1). DOI: 10.1190/geo2019-0435.1.
[20] ZHU Sen,WANG Yulin,XU Yunfan,et al. Asymmetric fiber grating overlapping spectrum demodulation technology based on convolutional network and wavelet transform noise reduction[J]. Optical Fiber Technology,2025,90. DOI: 10.1016/J.YOFTE.2025.104132.
[21] 董绍江,李洋,梁天,等. 基于CNN−BiLSTM的滚动轴承变工况故障诊断方法[J]. 振动. 测试与诊断,2022,42(5):1009-1016,1040. DONG Shaojiang,LI Yang,LIANG Tian,et al. Fault diagnosis method of rolling bearing based on CNN-BiLSTM under variable working conditions[J]. Journal of Vibration,Measurement & Diagnosis,2022,42(5):1009-1016,1040.
[22] 洪宇超,钱建固,叶源新,等. 基于时空关联特征的CNN−LSTM模型在基坑工程变形预测中的应用[J]. 岩土工程学报,2021,43(增刊2):108-111. HONG Yuchao,QIAN Jiangu,YE Yuanxin,et al. Application of CNN-LSTM model based on spatio-temporal correlation characteristics in deformation prediction of foundation pit engineering[J]. Chinese Journal of Geotechnical Engineering,2021,43(S2):108-111.
[23] 尹刚,朱淼,全鹏程,等. 基于PID搜索优化的CNN−LSTM−Attention铝电解槽电解温度预测方法研究[J]. 仪器仪表学报,2025,46(1):324-337. YIN Gang,ZHU Miao,QUAN Pengcheng,et al. Research on CNN-LSTM-Attention aluminum electrolyzer electrolysis temperature prediction method based on PID search optimization[J]. Chinese Journal of Scientific Instrument,2025,46(1):324-337.
[24] 张晗筱,王瑞琪,张亚婷. 基于CNN−LSTM的换热器污垢因子预测研究[J]. 化工学报,2025,76(4):1671-1679. ZHANG Hanxiao,WANG Ruiqi,ZHANG Yating. Prediction of scale factor of heat exchangers based on CNN-LSTM neural network[J]. CIESC Journal,2025,76(4):1671-1679.
[25] 朱仔旭,施成华,陈海勇,等. 考虑IPSO优化的CNN−LSTM盾构掘进参数混合预测模型[J]. 铁道科学与工程学报,2025,22(6):2690-2702. ZHU Zixu,SHI Chenghua,CHEN Haiyong,et al. Hybrid CNN-LSTM shield tunneling parameter prediction model considering IPSO optimization[J]. Journal of Railway Science and Engineering,2025,22(6):2690-2702.
[26] GUAN Wenhui,LI Binbin,XUE Shijie,et al. Research on CNN-attention regression prediction method based on GOA optimization[M]//Proceedings of the TEPEN International Workshop on Fault Diagnostic and Prognostic. Cham:Springer Nature Switzerland,2024:209-219.
[27] SU Ziyi,LIU Handong,QIAN Jinwu,et al. Hand gesture recognition based on sEMG signal and convolutional neural network[J]. International Journal of Pattern Recognition and Artificial Intelligence,2021,35(11). DOI: 10.1142/S0218001421510125.
[28] HE He,WANG Junhan,KOJIMA S,et al. Regression CNN based fast fading channel tracking using decision feedback channel estimation[J]. Journal of Signal Processing,2023,27(3):49-57. DOI: 10.2299/jsp.27.49
