Semi-supervised learning inversion of data derived from high-resolution electrical resistivity tomography based on forward modeling constraints
LI Guo-Hao1,2,3(), LYU Yu-Zeng1(), DONG Yi-Fan1, YU Hai-Tao2,3
1. College of Earth Sciences, Guilin University of Technology, Guilin 541004, China 2. Guangdong Provincial Academy of Building Research Group Co., Ltd., Guangzhou 510500, China 3. Guangdong Construction Engineering Quality & Safety Testing Head Station Co., Ltd., Guangzhou 510500, China
Supervised deep learning inversion relies on labels of subsurface media for training. However, the measured data from electrical resistivity tomography usually lack such labels. Supervised learning based on significant synthetic data generated through forward modeling fails to obtain reliable inversion results for complex data outside the training set. This study proposed a semi-supervised learning inversion method based on forward modeling constraints, combining with labeled and unlabeled data for training, to enhance the inversion accuracy and the capability to characterize complex geological structures. The inversion results of synthetic data demonstrate the strong generalization capability of the proposed method. The inversion results of measured data indicate that compared to conventional least-squares inversion, semi-supervised inversion provides higher resolution and particularly more accurate characterization of vertical lithological boundaries. Overall, the proposed method offers a novel approach for the inversion of the data derived from high-resolution electrical resistivity tomography.
Lyu Y Z, Cheng Y M, Lu G H, et al. Application of comprehensive geophysical prospecting technology in the exploration of sedimentary bauxite deposits in West Guangxi[J]. Mineral Resources and Geology, 2023, 37(5):1039-1045.
Zhang W, Zhou Y K, Liu L Y, et al. Application of the cross-borehole resistivity method in the monitoring of leakage for urban water supply pipelines[J]. Geophysical and Geochemical Exploration, 2024, 48(3):884-890.
Chen X Q, Li C G, Tian C J, et al. The application of high density electrical resistivity method to monitoring saltwater intrusion[J]. Geophysical and Geochemical Exploration, 2021, 45(5):1347-1353.
Jiang G Q, Hao S F, Yu Y X, et al. Landslide survey based on three-dimensional resistivity inversion:A case study of the Xuelang Mountain scenic spot,Wuxi,China[J]. Geophysical and Geochemical Exploration, 2024, 48(6):1720-1729.
Dai T Y, Xie S P. Application of resistivity technigue in the relic archaeology[J]. Progress in Geophysics, 2015, 30(6):2885-2891.
[6]
Monteiro S F A, Sultan S A, Represas P, et al. Joint inversion of gravity and geoelectrical data for groundwater and structural investigation:Application to the northwestern part of Sinai,Egypt[J]. Geophysical Journal International, 2006, 165(3):705-718.
Cheng B, Ding Y L, Xu Z F. Statistic modeling 2D inversion application in parallel resistivity survey lines[J]. Journal of Guilin University of Technology, 2016, 36(3):452-457.
Zhao T, Yu S J. GA-BP neural network algorithm-based nonlinear inversion for high density resistivity method[J]. Coal Geology & Exploration, 2017, 45(2):147-151.
[9]
郭谦. 基于先验约束的电阻率法探测深度学习反演方法[D]. 济南: 山东大学, 2019.
[9]
Guo Q. Learning inversion method of resistivity detection depth based on prior constraint[D]. Jinan: Shandong University, 2019.
Liu Z Y, Pang Y H, Zhang F K, et al. Deep learning resistivity inversion method based on multi-scale edge features[J]. Rock and Soil Mechanics, 2023, 44(11):3299-3306.
Song L, Yin X Y, Zong Z Y, et al. Deep learning seismic impedance inversion based on prior constraints[J]. Oil Geophysical Prospecting, 2021, 56(4):716-727,670.
Han H W, Liu H J, Sang W J, et al. Seismic and well logs integration for reservoir lateral porosity prediction based on semi-supervised learning[J]. Chinese Journal of Geophysics, 2022, 65(10):4073-4086.
Luo R Y, Chen J F, Yin X X. Review of the applications of machine learning in automatic detection of seismic events[J]. Progress in Geophysics, 2021, 36(3):923-932.
[14]
Laine S, Aila T M. Temporal ensembling for semi-supervised learning[EB/OL].2016:1610.02242.
[15]
Ronneberger O, Fischer P, Brox T. U-Net:Convolutional networks for biomedical image segmentation[C]// Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015. Cham: Springer International Publishing, 2015.
Gao H T, Sun N N, Sun K Y, et al. Comparative analysis of DnCNN and U-Net on suppression of seismic random noise[J]. Progress in Geophysics, 2021, 36(6):2441-2453.
[17]
Lou A G, Guan S Y, Loew M H. DC-UNet:Rethinking the U-Net architecture with dual channel efficient CNN for medical image segmentation[C]// Medical Imaging 2021:Image Processing.February 15-20, 2021. Online Only,USA.SPIE,2021:98.
[18]
Kingma D P, Ba J, Hammad M M. Adam: A method for stochastic optimization[EB/OL].2014:1412.6980.
[19]
Peng C, Gou W, Liu G. Ray-tracing-based input data selection RTM-A target oriented approach for clearer subsalt image[C]// Proceedings.June 1-4,2015.Madrid,Spain.EAGE Publications BV, 2015.
