Geology-informed intelligent prediction of aquifer groutability for water-inrush control in mining with an interpretable hybrid deep learning framework

Abstract

Accurate groutability prediction is essential not only for mine water-inrush prevention, but also for reducing excessive cement consumption and the associated carbon footprint of grouting operations. However, field geological datasets are often small, which limits the reliability and generalization of data-driven models. A geology-informed prediction framework is presented for predicting unit grouting amount (uga) from three routinely measured borehole variables: water inflow rate, hydraulic pressure, and groundwater level. A generative augmentation strategy was employed to expand the training data from 44 to 880 samples, and the input variables were organized into a physically informed feature sequence. Based on this representation, a hybrid deep-learning model integrating a bidirectional temporal convolutional network, a bidirectional gated recurrent unit, and an attention mechanism was developed, with its hyperparameters optimized using the Crested Porcupine Optimizer. The results demonstrate that data augmentation improved the test performance, with the coefficient of determination (R²) increasing from 0.8851 to 0.9293 and the root mean square error (RMSE) decreasing by 21.6%, effectively alleviating overfitting. External validation yielded R² values of 0.9982, 0.9884, and 0.9272 under identical, similar, and distinct geological settings, respectively. Distribution-shift analysis further highlighted the importance of hydrogeological-mechanism consistency for external generalization. Interpretability analysis using SHapley Additive exPlanations and Accumulated Local Effects indicated that borehole water inflow rate was the dominant predictor, contributing 75.4% to the model output, while high unit grouting amounts were associated with jointly elevated fracture connectivity and hydraulic pressure. The proposed framework provides practical support for accurate uga prediction and material-efficient grouting in intelligent mining.

Keywords

intelligent mining / mine water-inrush control / groutability prediction / explainable artificial intelligence / small-sample deep learning

References

[1] Kang HP, Li WZ, Gao FQ, Yang JW. Grouting theories and technologies for the reinforcement of fractured rocks surrounding deep roadways. Deep Undergr. Sci. Eng., 2023, 212

[2] P. Li, Q.S. Zhang, X. Zhang, S.C. Li, X.H. Li, and J.X. Zuo, Grouting diffusion characteristics in faults considering the interaction of multiple grouting, Int. J. Geomech., 17(2017), No. 5, art. No. 04016117.

[3] D. Ma, X.F. Gao, and J.X. Zhang, Co-exploitation of mine-derived geothermal energy: Recent advances and emerging perspectives, GeoEnergy Commun., 1(2025), No. 1, art. No. 12.

[4] H.Y. Duan, Y.S. Xie, L.C. Zou, B. Li, U. Håkansson, and V. Cvetkovic, Analysis of grout radial propagation in water-saturated rough-walled rock fractures, Int. J. Rock Mech. Min. Sci., 190(2025), art. No. 106101.

[5] Y.L. Zhe, K.P. Hou, Z.Y. Wang, S.F. Yang, and H.F. Sun, Prediction model for grouting volume using borehole image features and explainable artificial intelligence, Constr. Build. Mater., 470(2025), art. No. 140626.

[6] Z.X. Liu, S.N. Dong, H. Wang, and H.B. Shang, Mechanism and control of grout propagation in horizontal holes in fractured rock, Water, 14(2022), No. 24, art. No. 4062.

[7] Xiao F, Zhao ZY, Chen HM. A simplified model for predicting grout flow in fracture channels. Tunn. Undergr. Space Technol., 2017, 7011

[8] J.S. Liang, X.M. Du, H.Y. Fang, et al., Intelligent prediction model of a polymer fracture grouting effect based on a genetic algorithm-optimized back propagation neural network, Tunn. Undergr. Space Technol., 148(2024), art. No. 105781.

[9] Dai GL, Gong WM, Zhao XL, Zhou XQ. Static testing of pile-base post-grouting piles of the suramadu bridge. Geotech. Test. J., 2011, 34134

[10] T. Hu, Z.H. Wan, G.L. Dai, W.M. Gong, and B.W. Fang, Prediction and design of grouting parameters for long and large-diameter postgrouted drilled shafts, Int. J. Geomech., 24(2024), No. 2, art. No. 04023265.

[11] J.D. Niu, B. Wang, G.J. Chen, and K. Chen, Predicting of the unit grouting quantity in karst curtain grouting by the water permeability of rock strata, Appl. Sci., 9(2019), No. 22, art. No. 4814.

[12] Z.Y. Xu, A.L. Che, H.X. Zhou, Y.F. Shen, and W.G. He, Optimized deep learning modelling for predicting the diffusion range and state change of filling projects, Tunn. Undergr. Space Technol., 154(2024), art. No. 106073.

[13] Yang CP. Estimating cement take and grout efficiency on foundation improvement for Li–Yu–Tan dam. Eng. Geol., 2004, 7511

[14] S.H. Deng, X.L. Wang, Y.S. Zhu, F. Lv, and J.J. Wang, Hybrid grey wolf optimization algorithm-based support vector machine for groutability prediction of fractured rock mass, J. Comput. Civ. Eng., 33(2019), No. 2, art. No. 04018065.

[15] H.M. Zhu, M.Q. Huang, and Q.B. Zhang, TunGPR: Enhancing data-driven maintenance for tunnel linings through synthetic datasets, deep learning and BIM, Tunn. Undergr. Space Technol., 145(2024), art. No. 105568.

[16] Yin HC, Zhang GZ, Wu Q, et al. . Transfer learning with transformer-based models for mine water inrush prediction: A multivariate analysis using sparse and imbalanced monitoring data. Mine Water Environ., 2024, 434707

[17] S.H. Meng, Q. Wu, Y.F. Zeng, et al., Evaluating the impact of coal seam roof groundwater using variable weights theory: A special emphasis on skylight-type water inrush pattern, J. Hydrol. Reg. Stud., 56(2024), art. No. 102009.

[18] Chen ZW, Liu ZX, Chen JZ, Li XB, Huang LQ. Intelligent identification of acoustic emission Kaiser effect points and its application in efficiently acquiring in-situ stress. Int. J. Miner. Metall. Mater., 2025, 3271507

[19] S.X. Feng, Y.F. Zhao, Y.J. Wang, S.Y. Wang, and R.L. Cao, A comprehensive approach to karst identification and groutability evaluation–A case study of the Dehou reservoir, SW China, Eng. Geol., 269(2020), art. No. 105529.

[20] Nenchev B, Tao Q, Dong ZH, et al. . Evaluating data-driven algorithms for predicting mechanical properties with small datasets: A case study on gear steel hardenability. Int. J. Miner. Metall. Mater., 2022, 294836

[21] W.X. Li, J.T. Chen, J. Zhu, X.B. Ji, and Z.Q. Fu, Research on factor analysis and method for evaluating grouting effects using machine learning, Sci. Rep., 14(2024), art. No. 7782.

[22] X.Y. Li, Y. Pan, L.M. Zhang, and J.J. Chen, Dynamic and explainable deep learning-based risk prediction on adjacent building induced by deep excavation, Tunn. Undergr. Space Technol., 140(2023), art. No. 105243.

[23] W.J. Zhao, S.Y. Feng, J.X. Liu, and B.C. Sun, An explainable intelligent algorithm for the multiple performance prediction of cement-based grouting materials, Constr. Build. Mater., 366(2023), art. No. 130146.

[24] S.S. Lin, S.L. Shen, N. Zhang, and A.N. Zhou, Modelling the performance of EPB shield tunnelling using machine and deep learning algorithms, Geosci. Front., 12(2021), No. 5, art. No. 101177.

[25] Y. Zhou, S.Q. Li, C. Zhou, and H.B. Luo, Intelligent approach based on random forest for safety risk prediction of deep foundation pit in subway stations, J. Comput. Civ. Eng., 33(2019), No. 1, art. No. 05018004.

[26] M. Zounemat-Kermani and M. Kheimi, Explainable artificial intelligence in hydrology: A review, Water Resour. Manag., 40(2026), No. 3, art. No. 106.

[27] P. Bhasme and U. Bhatia, Improving the interpretability and predictive power of hydrological models: Applications for daily streamflow in managed and unmanaged catchments, J. Hydrol., 628(2024), art. No. 130421.

[28] H.H. Ding, X.M. Zhang, X.W. Chu, and Q. Wu, Simulation of groundwater dynamic response to hydrological factors in karst aquifer system, J. Hydrol., 587(2020), art. No. 124995.

[29] Zhao JH, Liu Q, Jiang CB, Wang DF. Application of rock mass index in the prediction of mine water inrush and grouting quantity. Geomech. Eng., 2022, 306503

[30] X.M. Liang, K. Feng, C.L. Wang, Z.Y. Hu, and C. He, Evaluation of rheology and washout resistance of cement-based grouts in water-rich strata, Tunn. Undergr. Space Technol., 157(2025), art. No. 106334.

[31] Z.Y. Wang, G.S. Zhao, Y. Zhou, and M.H. Ren, A numerical solution for unsteady permeation grouting of Bingham grout in saturated porous media considering the threshold pressure gradient, Comput. Geotech., 182(2025), art. No. 107138.

[32] B. Li, H. Wu, Q. Wu, Y.F. Zeng, and X.M. Guo, Prediction technology of mine water inflow based on entropy weight method and multiple nonlinear regression theory and its application, Geomech. Geophys. Geo Energy Geo Resour., 10(2024), No. 1, art. No. 127.

[33] Ramentol E, Caballero Y, Bello R, Herrera F. SMOTERSB*: A hybrid preprocessing approach based on over-sampling and undersampling for high imbalanced data-sets using SMOTE and rough sets theory. Knowl. Inf. Syst., 2012, 332245

[34] Zheng M, Li T, Zhu R, et al. . Conditional Wasserstein generative adversarial network-gradient penalty-based approach to alleviating imbalanced data classification. Inf. Sci., 2020, 5121009

[35] Mao XD, Li Q, Xie HR, Lau RYK, Wang Z, Smolley SP. On the effectiveness of least squares generative adversarial networks. IEEE Trans. Pattern Anal. Mach. Intell., 2019, 41122947

[36] Gao X, Deng F, Yue XH. Data augmentation in fault diagnosis based on the Wasserstein generative adversarial network with gradient penalty. Neurocomputing, 2020, 396487

[37] Bondielli A, Marcelloni F. A survey on fake news and rumour detection techniques. Inf. Sci., 2019, 49738

[38] E. Mahajan, H. Mahajan, and S. Kumar, EnsMulHateCyb: Multilingual hate speech and cyberbully detection in online social media, Expert Syst. Appl., 236(2024), art. No. 121228.

[39] Z.Y. Li and X.P. Xu, L2-BiTCN-CNN: Spatio-temporal features fusion-based multi-classification model for various internet applications identification, Comput. Netw., 243(2024), art. No. 110298.

[40] L. Yuan, Y.M. Ma, and Y.H. Liu, Ensemble deep learning models for protein secondary structure prediction using bidirectional temporal convolution and bidirectional long short-term memory, Front. Bioeng. Biotechnol., 11(2023), art. No. 1051268.

[41] Zhang YQ, Ma YM, Liu YH. Convolution-bidirectional temporal convolutional network for protein secondary structure prediction. IEEE Access, 2022, 10117469

[42] S. Akbar, Q. Zou, A. Raza, and F.K. Alarfaj, iAFPs-MvBiTCN: Predicting antifungal peptides using self-attention transformer embedding and transform evolutionary based multiview features with bidirectional temporal convolutional networks, Artif. Intell. Med., 151(2024), art. No. 102860.

[43] Chen JL, Jing HJ, Chang YH, Liu Q. Gated recurrent unit based recurrent neural network for remaining useful life prediction of nonlinear deterioration process. Reliab. Eng. Syst. Saf., 2019, 185372

[44] Deng YP, Wang L, Jia H, Tong XQ, Li F. A sequence-to-sequence deep learning architecture based on bidirectional GRU for type recognition and time location of combined power quality disturbance. IEEE Trans. Ind. Inform., 2019, 1584481

[45] M.A. Basset, R. Mohamed, and M. Abouhawwash, Crested Porcupine Optimizer: A new nature-inspired metaheuristic, Knowl. Based Syst., 284(2024), art. No. 111257.

[46] Lian GH, Liu XM, Wang Q, Shen CG, Wang Y, Mu WZ. Artificial intelligence-assisted non-metallic inclusion particle analysis in advanced steels using machine learning: A review. Int. J. Miner. Metall. Mater., 2026, 332401

[47] Qin X, Wang QH, Wang L, et al. . Interpretable machine learning-based stretch formability prediction of magnesium alloys. Int. J. Miner. Metall. Mater., 2025, 3281943

[48] Tang J, He MC, Qiao YF, Bian HB, Zhu C. Analytical solution for longitudinal responses of tunnels under combined effects of seismic waves and strike-slip faulting. J. Rock Mech. Geotech. Eng., 2026, 1821266

[49] J. Tang, M.C. He, H.B. Bian, and Y.F. Qiao, A novel numerical framework to simulate mechanical response and damage characteristics of mountain tunnels under faulting, Eng. Fail. Anal., 158(2024), art. No. 108030.

[50] J.X. Li, M. Wen, Z.M. Zhou, et al., Multi-objective optimization method for power supply and demand balance in new power systems, Int. J. Electr. Power Energy Syst., 161(2024), art. No. 110204.

[51] Aboalela R, Allehaibi KH, Alsaadi N, et al. . Harnessing feature pruning with optimal deep learning-based distributed denial of service cyberattack detection on IoT environment. Alex. Eng. J., 2025, 120584

RIGHTS & PERMISSIONS

University of Science and Technology Beijing