JOURNAL OF ROCK MECHANICS

JOURNAL OF ROCK MECHANICS

Multi-Class Prediction of Rock Mass Squeezing Severity in Tunnels Using Machine Learning Models

Document Type : Original Article

Authors
1 Department of Mining Engineering. Isfahan University of Technology. Isfahan 8415683111, Iran
2 Department of Mining Engineering. Hamedan University of Technology. Hamedan 65155579. Iran
3 Department of Mining Engineering, Isfahan University of Technology, Isfahan, Iran.
Abstract
Rock mass squeezing is one of the main challenges in tunnel excavation through weak rock masses under high in-situ stress conditions, and it can cause significant deformation, increased pressure on the support system, reduced advance rate, and tunnel instability. In this study, two machine learning models, namely a decision tree (J48) and a logistic model tree (LMT), were developed to predict squeezing severity across five classes. The database consisted of 139 case histories collected from tunneling projects in China, Nepal, India, Austria, Japan, and Turkey. Five parameters, namely overburden depth, tunnel diameter, support system stiffness, stress-to-strength ratio, and rock mass basic quality index, were selected as input variables. To address class imbalance, the SMOTE technique was applied, increasing the number of samples to 190, with 38 samples per class. Model performance was evaluated using 10-fold cross-validation and the metrics of accuracy, sensitivity, precision, Matthews correlation coefficient, Kappa, and AUC. The results showed that the logistic model tree, with an accuracy of 84.21% and an AUC of 0.96, outperformed the decision tree, which achieved an accuracy of 82.10% and an AUC of 0.94. Nevertheless, the simpler structure of the decision tree enhances its interpretability and practical applicability. Both models demonstrated acceptable capability for multi-class prediction of squeezing severity and may serve as a supplementary tool for preliminary assessment of squeezing risk
Keywords
Subjects

[1]    G. Barla, Squeezing rocks in tunnels,  ISRM News Journal. 2 (1995) 44–49.
[2]    E. Hoek, P. Marinos, Predicting tunnel squeezing problems in weak heterogeneous rock masses, Tunnels Tunnelling International. 32 (2000) 45–51.
[3]    R. Jimenez, D. Recio, A linear classifier for probabilistic prediction of squeezing conditions in Himalayan tunnels, Engineering Geology. 121 (2011) 101–109.
[4]    C. Cao, C. Shi, M. Lei, W. Yang, J. Liu, Squeezing failure of tunnels: A case study, Tunnelling and Underground Space Technology. 77 (2018) 188–203. https://doi.org/10.1016/J.TUST.2018.04.007.
[5]    G. Gioda, A. Cividini, Numerical methods for the analysis of tunnel performance in squeezing rocks, Rock Mechanics and Rock Engineering. 29 (1996) 171–193. https://doi.org/10.1007/BF01042531.
[6]    T.D. O’Rourke, Guidelines for tunnel lining design, in: ASCE, 1984.
[7]    J.L. Jethwa, B. Singh, B. Singh, Estimation of ultimate rock pressure for tunnel linings under squeezing rock conditions—a new approach, in: Design and Performance of Underground Excavations: ISRM Symposium—Cambridge, UK, 3–6 September 1984, Thomas Telford Publishing, 1984: pp. 231–238.
[8]    E. Ghasemi, H. Gholizadeh, Development of Two Empirical Correlations for Tunnel Squeezing Prediction Using Binary Logistic Regression and Linear Discriminant Analysis, Geotechnical and Geological Engineering. 37 (2018) 3435–3446. https://doi.org/10.1007/S10706-018-00758-0.
[9]    E.T. Brown, E. Hoek, Underground excavations in rock, CRC Press, 1980.
[10]    C. Carranza-Torres, Elasto-plastic solution of tunnel problems using the generalized form of the hoek-brown failure criterion, International Journal of Rock Mechanics and Mining Sciences. 41 (2004) 629–639. https://doi.org/10.1016/J.IJRMMS.2004.03.111.
[11]    M. Liang, H. Peng, W. Xie, B. Yu, Y. Han, M. Zhu, G. Song, N. Huang, Dynamic multiclass prediction of tunnel squeezing intensity with stacking model and Markov process, Tunnelling and Underground Space Technology. 146 (2024) 105632. https://doi.org/10.1016/J.TUST.2024.105632.
[12]    E. Ghasemi, S. Hassani, M.H. Kadkhodaei, M. Bahri, R. Romero-Hernandez, E.J. Mascort-Albea, An Intelligent Approach to Predict the Squeezing Severity and Tunnel Deformation in Squeezing Grounds, Transportation Infrastructure Geotechnology 2024 11:6. 11 (2024) 3992–4016. https://doi.org/10.1007/S40515-024-00434-2.
[13]    A. Shafiei, H. Parsaei, M.B. Dusseault, Rock squeezing prediction by a support vector machine classifier, in: 46th US Rock Mechanics/Geomechanics Symposium, American Rock Mechanics Association, 2012.
[14]    X. Feng, R. Jimenez, Predicting tunnel squeezing with incomplete data using Bayesian networks, Engineering Geology. 195 (2015) 214–224.
[15]    Y. Sun, X. Feng, L. Yang, Predicting Tunnel Squeezing Using Multiclass Support Vector Machines, Advances in Civil Engineering. 2018 (2018).
[16]    E. Ghasemi, H. Gholizadeh, Prediction of squeezing potential in tunneling projects using data mining-based techniques, Geotechnical Geological Engineering. (2018) 1–10. https://doi.org/10.1007/s10706-018-0705-6.
[17]    E. Ghasemi, H. Gholizadeh, Development of Two Empirical Correlations for Tunnel Squeezing Prediction Using Binary Logistic Regression and Linear Discriminant Analysis, Geotechnical and Geological Engineering. 37 (2019) 3435–3446. https://doi.org/10.1007/s10706-018-00758-0.
[18]    J. Zhang, D. Li, Y. Wang, Predicting tunnel squeezing using a hybrid classifier ensemble with incomplete data, Bulletin of Engineering Geology and The Environment. (2020) 1–12.
[19]    Y. Chen, T. Li, P. Zeng, J. Ma, E. Patelli, B. Edwards, Dynamic and Probabilistic Multi-class Prediction of Tunnel Squeezing Intensity, Rock Mechanics and Rock Engineering. (2020) 1–22.
[20]    Z. Huang, M. Liao, H. Zhang, J. Zhang, S. Ma, Q. Zhu, Predicting Tunnel Squeezing Using the SVM-BP Combination Model, Geotechnical and Geological Engineering. 40 (2022) 1387–1405. 
[21]    H. Fathipour-Azar, Multi-level Machine Learning-Driven Tunnel Squeezing Prediction: Review and New Insights, Archives of Computational Methods in Engineering. 29 (2022) 5493–5509. 
[22]    J. Zhou, S. Zhu, Y. Qiu, D.J. Armaghani, A. Zhou, W. Yong, Predicting tunnel squeezing using support vector machine optimized by whale optimization algorithm, Acta Geotechnica. 17 (2022) 1343–1366.
[23]    Y. Bo, X. Huang, Y. Pan, Y. Feng, P. Deng, F. Gao, P. Liu, Q. Liu, Robust model for tunnel squeezing using Bayesian optimized classifiers with partially missing database, Underground Space. 10 (2023) 91–117. https://doi.org/10.1016/J.UNDSP.2022.11.001.
[24]    M.H. Kadkhodaei, E. Ghasemi, S. Mahdavi, Modelling tunnel squeezing using gene expression programming: a case study, Proceedings of the Institution of Civil Engineers - Geotechnical Engineering. (2023) 1–30. https://doi.org/10.1680/JGEEN.22.00151.
[25]    X. Geng, S. Wu, Y. Zhang, J. Sun, H. Cheng, Z. Zhang, S. Pu, Developing hybrid XGBoost model integrated with entropy weight and Bayesian optimization for predicting tunnel squeezing intensity, Natural Hazards. 119 (2023) 751–771. https://doi.org/10.1007/S11069-023-06137-0/METRICS.
[26]    M.H. Kadkhodaei, V. Amirkiyaei, E. Ghasemi, Developing two robust hybrid models for predicting tunnel deformation in squeezing prone grounds, Transportation Geotechnics. 45 (2024) 101231. https://doi.org/10.1016/J.TRGEO.2024.101231.
[27]    D. Liang, Z. Rui, F. Yuguang, A robust evaluating strategy of tunnel deterioration using ensemble machine learning algorithms, Engineering Applications of Artificial Intelligence. 133 (2024) 108364. https://doi.org/10.1016/J.ENGAPPAI.2024.108364.
[28]    P. Guan, G. Ou, F. Liang, W. Luo, Q. Wang, C. Pei, X. Che, Tunnel squeezing prediction based on partially missing dataset and optimized machine learning models, Frontiers in Earth Science. 13 (2025) 1511413. https://doi.org/10.3389/FEART.2025.1511413/TEXT.
[29]    J. YY, O. GZ, W. H, Z. GH, Prediction of tunnel squeezing based on evidence theory, Journal of Basic Science and Engineering. 29 (2021) 1156–1170. https://doi.org/10.16058/j.issn.1005-0930.2021.05.007.
[30]    N. V. Chawla, K.W. Bowyer, L.O. Hall, W.P. Kegelmeyer, SMOTE: Synthetic Minority Over-sampling Technique, Journal of Artificial Intelligence Research. 16 (2002) 321–357. https://doi.org/10.1613/JAIR.953.
[31]    A. Fernández, S. García, F. Herrera, N. V. Chawla, SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary, Journal of Artificial Intelligence Research. 61 (2018) 863–905. https://doi.org/10.1613/JAIR.1.11192.
[32]    M.H. Kadkhodaei, E. Ghasemi, Development of a Semi-quantitative Framework to Assess Rockburst Risk Using Risk Matrix and Logistic Model Tree, Geotechnical and Geological Engineering 2022 40:7. 40 (2022) 3669–3685. https://doi.org/10.1007/S10706-022-02122-9.
[33]    N. Landwehr, M. Hall, E. Frank, Logistic model trees, Machine Learning. 59 (2005) 161–205. https://doi.org/10.1007/s10994-005-0466-3.
[34]    W. Chen, X. Xie, J. Wang, B. Pradhan, H. Hong, D.T. Bui, Z. Duan, J. Ma, A comparative study of logistic model tree, random forest, and classification and regression tree models for spatial prediction of landslide susceptibility, CATENA. 151 (2017) 147–160. https://doi.org/10.1016/J.CATENA.2016.11.032.
[35]    T.T. Wong, Performance evaluation of classification algorithms by k-fold and leave-one-out cross validation, Pattern Recognition. 48 (2015) 2839–2846. https://doi.org/10.1016/J.PATCOG.2015.03.009.
[36]    E. Kurniawan, F. Nhita, A. Aditsania, D. Saepudin, C5.0 algorithm and synthetic minority oversampling technique (SMOTE) for rainfall forecasting in bandung regency, 2019 7th International Conference on Information and Communication Technology, ICoICT 2019. (2019). https://doi.org/10.1109/ICOICT.2019.8835324.