To address the limitations of traditional empirical interpretation methods for advanced ground penetrating radar (GPR) detection images of tunnels, including inconsistent standards, low efficiency, and high rates of misinterpretation and missed detection, this study proposed an intelligent inversion method for radar B-scan waveform images of adverse geology based on an improved generative adversarial network (GAN). A multi-condition synthetic dataset was constructed through electromagnetic numerical simulation combined with random generation constraints of medium parameters, incorporating three types of geological anomalies (irregular cavities, fracture zones, and cracks) and their combinations. The GAN generator was optimized by integrating Unet convolutional neural network architecture and dilated convolution modules, and structural similarity (SSIM) loss function was incorporated into the loss function to enhance the extraction capability of waveform image features and model inversion performance. Training evaluation results demonstrate that the improved GAN model achieves approximately 4% and 10% higher inversion accuracy than conventional GAN and standalone Unet models, respectively. An inversion experiment using actual engineering data from fracture zone prediction was carried out. Comparative analysis with digital drilling and excavation verification data confirms that the inversion prediction results of the proposed method are basically consistent with actual conditions, and it can provide reliable geological forecasting for tunnel excavation and support.
LiShucai, LiuBin, SunHuaifeng, et al. State of art and trends of advanced geological prediction in tunnel construction[J]. Chinese Journal of Rock Mechanics and Engineering, 2014, 33(6): 1090-1113.
Editorial Department of China Journal of Highway and Transport. Review on China’s traffic tunnel engineering research: 2022[J]. China Journal of Highway and Transport, 2022, 35(4): 1-40.
ZhangKunmu, PengHao, LiangMing, et al. Quantitative evaluation of tunnel rock mass integrity based on MWD technology[J]. Journal of China & Foreign Highway, 2025, 45(6): 245-253.
ZhengJianlong, ChenMengjie, LiuChaochao. Review of intelligent construction development of highway infrastructures and its prospects[J]. Journal of China & Foreign Highway, 2025, 45(2): 1-20.
WangMinling, LiangZhihuan, WangHonghua, et al. Review of reverse time migration in ground penetrating radar[J]. Progress in Geophysics, 2019, 34(5): 2087-2096.
ZhangChongmin, ZhangFengkai, LiYao. Study of full waveform inversion of advance tunnel geological prediction by ground penetrating radar[J]. Tunnel Construction, 2019, 39(1): 102-109.
[13]
SollaM, Pérez-graciaV, FontulS. A review of GPR application on transport infrastructures: Troubleshooting and best practices[J]. Remote Sensing, 2021, 13(4): 672.
[14]
AlvarezJ K, KodagodaS. Application of deep learning image-to-image transformation networks to GPR radargrams for sub-surface imaging in infrastructure monitoring[C]//2018 13th IEEE Conference on Industrial Electronics and Applications (ICIEA). Wuhan, China. IEEE, 2018: 611-616.
[15]
XieL H, ZhaoQ, MaC G, et al. U-net: Deep-learning schemes for ground penetrating radar data inversion[J]. Journal of Environmental and Engineering Geophysics, 2020, 25(2): 287-292.
[16]
LiuB, RenY X, LiuH C, et al. GPRInvNet: Deep learning-based ground-penetrating radar data inversion for tunnel linings[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(10): 8305-8325.
[17]
WangJ, LiuH C, JiangP, et al. GPRI2Net: A deep-neural-network-based ground penetrating radar data inversion and object identification framework for consecutive and long survey lines[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5106320.
[18]
JiY T, ZhangF K, WangJ, et al. Deep neural network-based permittivity inversions for ground penetrating radar data[J]. IEEE Sensors Journal, 2021, 21(6): 8172-8183.
WuMingze, LiuQinghua, OuyangShan. Two-stage GPR image inversion method based on generative adversarial network[J]. Chinese Journal of Radio Science, 2025, 40(1): 141-154.
LiuDongkun, JuNengpan, HuoYuxiang. Analysis of the spectrum difference of ground penetrating radar (GPR) for different media fillings[J]. Modern Tunnelling Technology, 2013, 50(5): 23-28.
SunHuaifeng, ChengMing, WuQilong, et al. A multi-scale grid scheme in three-dimensional transient electromagnetic modeling using FDTD[J]. Chinese Journal of Geophysics, 2018, 61(12): 5096-5104.
[27]
GoodfellowI J, Pouget-AbadieJ, MirzaM, et al. Generative adversarial nets[C]. Proceedings of the 27th International Conference on Neural Information Processing Systems, Montreal, Canada, 2014.
[28]
WangY Z, QinH, MiaoF. A multi-path encoder network for GPR data inversion to improve defect detection in reinforced concrete[J]. Remote Sensing, 2022, 14(22): 5871.
LiShinian, WangXiurong, LinTian, et al. Numerical simulation of 3D ground penetrating radar based on GprMax for the road cavity[J]. The Chinese Journal of Geological Hazard and Control, 2020, 31(3): 132-138.
LiRongpeng, SongXueli, YiWen, et al. A joint robust sparse Bayesian learning model with the mixture of Gaussians for the structural damage identificatio[J]. China Journal of Highway and Transport, 2024(5): 300-313.
TianGang, ZhangHaijun, WangChengliang, et al. Research on identification of subsurface defects of urban roads and collapse risk assessment[J]. Journal of China & Foreign Highway, 2025, 45(2): 46-54.
[35]
GaoR. Rethinking dilated convolution for real-time semantic segmentation[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Vancouver, BC, Canada. IEEE, 2023: 4675-4684.
[36]
张志宇, 李新利, 杨国田. 基于改进生成对抗网络pix2pix的火电厂NO x 排放预测[J]. 中国电机工程学报, 2023, 43(12): 4678-4687.
[37]
ZhangZhiyu, LiXinli, YangGuotian. NO x emission prediction of thermal power plant based on improved generative adversarial network pix2pix[J]. Proceedings of the CSEE, 2023, 43(12): 4678-4687.