Geomechanics and Engineering

  • 제39권3호
  • /
  • Pages.241-256
  • /
  • 2024
  • /
  • 2005-307X(pISSN)
  • /
  • 2092-6219(eISSN)
테크노프레스 (Techno-Press)
권호 표지
DOI QR코드

DOI QR Code

The effect of soil physical properties on predicting shear strength parameters based on comparing ensemble learning, deep learning, and support vector machine models

  • Ba-Quang-Vinh Nguyen (School of Civil Engineering and Management, International University) ;
  • Yun-Tae Kim (Ocean Engineering Department, Pukyong National University)
  • 투고 : 2023.11.03
  • 심사 : 2024.10.14
  • 발행 : 2024.11.10
⟨ 이전 논문 다음 논문 ⟩

초록

The shear strength (SS) of soil is a critical parameter utilized in the design of civil engineering projects. The SS parameters, including cohesion (c) and friction angle (𝜑), can be determined through methods conducted either in the field or within a laboratory environment. However, the traditional method for determining SS parameters are not only costly but also time-consuming. Recently, the application of machine learning (ML) in geotechnical problems has received increasing attention. In order to select an appropriate ML model and assess the effect of physical properties on the SS of soil. This research endeavors to predict critical SS parameters of soil through the application of five machine learning (ML) models, integrating easily-available physical soil index, including specific gravity (G), saturation degree (Sr), liquid limit (LL), silt content (SC), and clay content (CC). The used ML techniques include Extreme Gradient Boosting (XGBoost), Random Forest (RF), Multilayer Perceptron (MLP), Support Vector Machine (SVM), and Convolutional Neural Network (CNN). A range of metrics, encompassing the root mean square error (RMSE), mean absolute error (MAE), and determination coefficient (R2) were used to measure the predictive efficacy of the employed models as well as compare the performance of the used ML models. The values of R2 range from 0.769 to 0.987 indicate that all ML models exhibit excellent predictive capabilities for estimating SS parameters, in which the XGBoost, and CNN techniques show outperforming results compared to the other models. The study uses decision tree feature importance (DTFI) and coefficient feature importance (CFI) techniques to investigate how various physical properties impact the predictive capabilities of the model and indicates that both G and LL have a substantial impact on the predictive accuracy of cohesion and friction angle.

키워드

과제정보

This research is funded by Vietnam National University Ho Chi Minh City (VNU-HCM) under grant number C2023-28-07.

참고문헌

  1. Adunoye, G.O. (2014a), "Fines content and angle of internal friction of a lateritic soil: an experimental study", Am. J. Eng. Res., 3(3), 16-21.
  2. Adunoye, G.O. (2014b), "Study of relationship between fines content and cohesion of soil", British J. Appl. Sci. Technol., 4(4), 682-692. https://doi.org/10.9734/BJAST/2014/6290.
  3. Amin Benbouras, M. and Petrisor, A.I. (2021), "Prediction of swelling index using advanced machine learning techniques for cohesive soils", Appl. Sci., 11(2), 536. https://doi.org/10.3390/app11020536.
  4. Amjad, M., Ahmad, I., Ahmad, M., Wroblewski, P., Kaminski, P., and Amjad, U. (2022), "Prediction of pile bearing capacity using XGBoost algorithm: Modeling and performance evaluation", Appl. Sci., 12(4), 2126. https://doi.org/10.3390/app12042126.
  5. Arabameri, A., Saha, S., Roy, J., Chen, W., Blaschke, T. and Tien Bui, D. (2020), "Landslide susceptibility evaluation and management using different machine learning methods in the Gallicash River Watershed, Iran", Remote Sens., 12(3), 475. https://doi.org/10.3390/rs12030475.
  6. Ates, E.C., Bostanci, E. and Guzel, M.S. (2021), "Comparative performance of machine learning algorithms in cyberbullying detection: Using turkish language preprocessing techniques", ArXiv:2101.12718. https://doi.org/10.48550/arXiv.2101.12718.
  7. Bandara, A., Hettiarachchi, Y., Hettiarachchi, K., Munasinghe, S., Wijesinghe, I. and Thayasivam, U. (2020), "A generalized ensemble machine learning approach for landslide susceptibility modeling", Data Management, Analytics and Innovation: Proceedings of ICDMAI 2019, 2, 71-93.
  8. Benemaran, R.S. and Esmaeili-Falak, M. (2023), "Predicting the Young's modulus of frozen sand using machine learning approaches: State-of-the-art review", Geomech. Eng., 34(5), 507-527. https://doi.org/10.12989/gae.2023.34.5.507.
  9. Blahova, K., Sevelova, L. and Pilarova, P. (2013), "Influence of water content on the shear strength parameters of clayey soil in relation to stability analysis of a hillside in Brno region", Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, 61(6), 1583-1588. https://doi.org/10.11118/actaun201361061583.
  10. Brakorenko, N., Leonova, A. and Nikitenkov, A. (2019), "Effect of soil water saturation on slope stability: Tomsk case study", E3S Web of Conferences, 98, 5005.
  11. Brownlee, J. (2016), Deep learning with python: Develop deep learning models on theano and tensorflow using keras. Machine Learning Mastery.
  12. D.T., Lofman, O., Revhaug, I. and Dick, O. (2011), "Landslide susceptibility analysis in the Hoa Binh province of Vietnam using statistical index and logistic regression", Nat. Hazards, 59(3), 1413. https://doi.org/10.1007/s11069-011-9844-2.
  13. Chaganti, S.Y., Nanda, I., Pandi, K.R., Prudhvith, T.G. and Kumar, N. (2020), "Image Classification using SVM and CNN", Proceedings of the 2020 International Conference on Computer Science, Engineering and Applications (ICCSEA).
  14. Chao, Z., Fowmes, G. and Dassanayake, S.M. (2021), "Comparative study of hybrid artificial intelligence approaches for predicting peak shear strength along soil-geocomposite drainage layer interfaces", Int. J. Geosynth. Ground Eng., 7(3), 60. https://doi.org/10.1007/s40891-021-00299-2.
  15. Chao, Z., Shi, D., Fowmes, G., Xu, X., Yue, W., Cui, P., Hu, T., and Yang, C. (2023), "Artificial intelligence algorithms for predicting peak shear strength of clayey soil-geomembrane interfaces and experimental validation", Geotext. Geomembranes, 51(1), 179-198. https://doi.org/10.1016/j.geotexmem.2022.10.007.
  16. Chen, T. and Guestrin, C. (2016). "Xgboost: A scalable tree boosting system", Proceedings of the 22nd Acm Sigkdd International Conference on Knowledge Discovery and Data Mining.
  17. Chen, W., Zhang, S., Li, R. and Shahabi, H. (2018). "Performance evaluation of the GIS-based data mining techniques of best-first decision tree, random forest, and naive Bayes tree for landslide susceptibility modeling", Sci. Total Environ., 644, 1006-1018. https://doi.org/10.1016/j.scitotenv.2018.06.389.
  18. Chen, X. and Chen, W. (2021). "GIS-based landslide susceptibility assessment using optimized hybrid machine learning methods", Catena, 196, 104833. https://doi.org/10.1016/j.catena.2020.104833.
  19. Cheng, Y.S., Yu, T.T. and Son, N.T. (2021), "Random forests for landslide prediction in tsengwen river watershed, central taiwan", Remote Sens., 13(2), 199. https://doi.org/10.3390/rs13020199.
  20. Chitra, R. and Gupta, M. (2014), "Neural networks for assessing shear strength of soils", Int. J. Recent Dev. Eng. Technol., 3(4), 24-32.
  21. Cortes, C. and Vapnik, V. (1995), "Support-vector networks", Machine Learning, 20, 273-297.
  22. Dadkhah, R., Ghafoori, M., Ajalloeian, R. and Lashkaripour, G.R. (2010), "The effect of scale direct shear tests on the strength parameters of clayey sand in Isfahan city, Iran", J. Appl. Sci., 18. DOI: 10.3923/jas.2010.2027.2033.
  23. Ersoy, H., Karsli, M.B., Cellek, S., Kul, B., Baykan, I. and Parsons, R.L. (2013). "Estimation of the soil strength parameters in Tertiary volcanic regolith (NE Turkey) using analytical hierarchy process", J. Earth Syst. Sci., 122, 1545-1555. https://doi.org/10.1007/s12040-013-0366-z
  24. Esmaeili-Falak, M. and Benemaran, R.S. (2023), "Ensemble deep learning-based models to predict the resilient modulus of modified base materials subjected to wet-dry cycles", Geomech. Eng., 32(6), 583-600. https://doi.org/10.12989/gae.2023.32.6.583.
  25. Fattahi, H. and Hasanipanah, M. (2021). "Predicting the shear strength parameters of rock: A comprehensive intelligent approach", Geomech. Eng., 27(5), 511. https://doi.org/10.12989/gae.2021.27.5.511.
  26. Garcia, A.J.H., Jaime, Y.N.M., Contreras, A.M.Z., Bastardo, L.D. S. and Llovera, F.A.S. (2012). "Savanna soil water content effect on its shear strength-compaction relationship", Revista Cientifica UDO Agricola, 12(2), 324-337.
  27. Girshick, R. (2015), "Fast r-cnn. Proceedings of the IEEE International Conference on Computer Vision, 1440-1448.
  28. Goktepe, A.B., Altun, S., Altintas, G. and Tan, O. (2008), "Shear strength estimation of plastic clays with statistical and neural approaches", Build. Environ., 43(5), 849-860. https://doi.org/10.1016/j.buildenv.2007.01.022.
  29. Guo, T., Dong, J., Li, H. and Gao, Y. (2017), "Simple convolutional neural network on image classification", Proceedings of the 2017 IEEE 2nd International Conference on Big Data Analysis (ICBDA).
  30. Hosseini, M., Movahedi Naeini, S.A.R., Dehghani, A.A. and Zeraatpisheh, M. (2018), "Modeling of soil mechanical resistance using intelligent methods", J. Soil Sci. Plant Nutrition, 18(4), 939-951. https://doi.org/10.4067/S0718-95162018005002702.
  31. Inan, M.S.K. and Rahman, I. (2023), "Explainable AI integrated feature selection for landslide susceptibility mapping using TreeSHAP", SN Comput. Sci., 4(5), 482. https://doi.org/10.1007/s42979-023-01960-5.
  32. Iyeke, S.D., Eze, E.O., Ehiorobo, J.O. and Osuji, S.O. (2016), "Estimation of shear strength parameters of lateritic soils using artificial neural network", Nigerian J. Technol., 35(2), 260-269. https://doi.org/10.4314/njt.v35i2.5.
  33. Jain, R., Jain, P.K. and Bhadauria, S.S. (2010), "Computational approach to predict soil shear strength", Int. J. Eng. Sci. Tech., 2(8), 3874-3885.
  34. Karamizadeh, S., Abdullah, S.M., Halimi, M., Shayan, J. and javad Rajabi, M. (2014), "Advantage and drawback of support vector machine functionality", Proceedings of the 2014 International Conference on Computer, Communications, and Control Technology (I4CT).
  35. Kayabali, O., Akturk, M., Fener, A.B., Ustun, O., Dikmen, O. and Harputlugil, F.H. (2015), "Evaluation of undrained shear strength of fine-grained soils in consideration of soil plasticity", Yerbilimleri, 36(3), 121-136. https://doi.org/10.17824/yrb.99784.
  36. Khanlari, G.R., Heidari, M., Momeni, A.A. and Abdilor, Y. (2012), "Prediction of shear strength parameters of soils using artificial neural networks and multivariate regression methods", Eng. Geol., 131, 11-18. https://doi.org/10.1016/j.enggeo.2011.12.006.
  37. Kulsoom, I., Hua, W., Hussain, S., Chen, Q., Khan, G. and Shihao, D. (2023), "SBAS-InSAR based validated landslide susceptibility mapping along the Karakoram Highway: A case study of Gilgit-Baltistan, Pakistan", Scientific Reports, 13(1), 3344. https://doi.org/10.1038/s41598-023-30009-z.
  38. Kumari, D. (2009), A study on the effect of moisture on strength characteristics of river sand.
  39. Kwak, N.S. and Ko, T.Y. (2022), "Machine learning-based regression analysis for estimating Cerchar abrasivity index", Geomech. Eng., 29(3), 219-228. https://doi.org/10.12989/gae.2022.29.3.219.
  40. Lade, P.V., Liggio, C.D. and Yamamuro, J.A. (1998), "Effects of non-plastic fines on minimum and maximum void ratios of sand", Geotech. Test. J., 21, 336-347. https://doi.org/10.1520/GTJ11373J.
  41. Langfelder, L.J. and Nivargikar, V.R. (1967), "Some factors influencing shear strength and compressibility of compacted soils", Highway Res. Record, 177.
  42. LeCun, Y., Bengio, Y. and Hinton, G. (2015), "Deep learning", Nature, 521(7553), 436.
  43. Lee, C.Y. and Chern, S.G. (2013), "Application of a support vector machine for liquefaction assessment", J. Mar. Sci. Tech., 21(3), https://doi.org/10.6119/JMST-012-0518-3.
  44. Lei, X., Pan, H. and Huang, X. (2019), "A dilated CNN model for image classification", IEEE Access, 7, 124087-124095. https://doi.org/10.1109/ACCESS.2019.2927169.
  45. Li, J., Geng, H., Zhang, G. and Jin, G. (2021), "Influence of silt content on shear strength of sandy soil", IOP Conference Series: Earth and Environmental Science, 768(1), 12088.
  46. Mohammadi, M., Fatemi Aghda, S.M., Talkhablou, M. andCheshomi, A. (2022), "Prediction of the shear strength parameters from easily-available soil properties by means of multivariate regression and artificial neural network methods", Geomech. Geoeng., 17(2), 442-454. https://doi.org/10.1080/17486025.2020.1778194.
  47. Mollahasani, A., Alavi, A.H., Gandomi, A.H. and Rashed, A. (2011), "Nonlinear neural-based modeling of soil cohesion intercept", KSCE J. Civil Eng., 15, 831-840. https://doi.org/10.1007/s12205-011-1154-4.
  48. Mousavi, S.M., Alavi, A.H., Gandomi, A.H. and Mollahasani, A. (2011), "Nonlinear genetic-based simulation of soil shear strength parameters", J. Earth Syst. Sci., 120, 1001-1022. https://doi.org/10.1007/s12040-011-0119-9.
  49. Mujtaba, H. (2014), "Effect of void ratio and gradation on shear strength parameters of granular soils", Pakistan J. Sci., 66(3).
  50.  Murthy, V.N.S. (2002), Geotechnical engineering: principles and practices of soil mechanics and foundation engineering, CRC press.
  51. Nguyen, B.Q.V. and Kim, Y.T. (2021), "Landslide spatial probability prediction: a comparative assessment of naive Bayes, ensemble learning, and deep learning approaches", Bull. Eng. Geol. Environ., 80, 4291-4321. https://doi.org/10.1007/s10064-021-02194-6.
  52. Nguyen, B.Q.V., Song, C.H. and Kim, Y.T. (2022), "A hybrid physical and machine learning model for assessing landslide spatial probability caused by raising of ground water table and earthquake in Atsuma, Japan-case study", KSCE J. Civil Eng., 26(8), 3416-3429. https://doi.org/10.1007/s12205-022-1656-2.
  53. Ojuri, O.O. (2013), "Predictive shear strength models for tropical lateritic soils", J. Eng., 2013. https://doi.org/10.1155/2013/595626.
  54. Palani, K.A., Maurya, S., Gupta, M. and Chitra, R. (2020), "Degree of initial saturation and shear strength of soil in a triaxial loading", Int. J. Emerg. Technol. Innov. Res., 7(4), 805-812.
  55. Pereira, L., Godinho, L. and Branco, F.G. (2023), "Predicting unconfined compression strength and split tensile strength of soil-cement via artificial neural networks", Geomech. Eng., 33(6), 611-624. https://doi.org/10.12989/gae.2023.33.6.611.
  56. Pham, B.T., Hoang, T.A., Nguyen, D.M. and Bui, D.T. (2018), "Prediction of shear strength of soft soil using machine learning methods", Catena, 166, 181-191. https://doi.org/10.1016/j.catena.2018.04.004.
  57. Probst, P., Wright, M.N. and Boulesteix, A. (2019), "Hyperparameters and tuning strategies for random forest", Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 9(3), e1301.
  58. Quang, V.N.B., Viet, L.D., Chi, C.N., Duc, P.V.N. and Quang, B. N. (2021), "Predicting landslide spatial probability in Quang Ngai, Vietnam using deep learning technique", Proceedings of the 4th Asia Pacific Meeting on Near Surface Geoscience & Engineering, 2021(1), 1-5.
  59. Roman, M.J. (2014), Correlation between shear strength at plastic limit and liquid limit Incorporating electrical resistivity.
  60. Roy, S. and Dass, G. (2014), "Statistical models for the prediction of shear strength parameters at Sirsa, India", Int. J. Civil Struct. Eng., 4(4), 483.
  61. Roy, S., Prajapati, A.K. and Maurya, A.K. (2019), "Prediction of shear strength parameters using multiple regression analysis", Int. J. Landscape Plann. Architect., 5(2), 26-39.
  62. Sahin, E.K. (2020), "Assessing the predictive capability of ensemble tree methods for landslide susceptibility mapping using XGBoost, gradient boosting machine, and random forest", SN Appl. Sci., 2(7), 1308. https://doi.org/10.1007/s42452-020-3060-1.
  63. Salari, P., Lashkaripour, G. and Ghafoori, M. (2015). "Presentation of empirical equations for estimating internal friction angle of SP and SC soils in Mashhad, Iran using standard penetration and direct shear tests and comparison with previous equations", Int. J. Geograph. Geol., 4(5), 89. https://doi.org/10.4236/ojg.2015.55021.
  64. Shahani, N. M., Kamran, M., Zheng, X., Liu, C. and Guo, X. (2021). "Application of gradient boosting machine learning algorithms to predict uniaxial compressive strength of soft sedimentary rocks at Thar Coalfield", Adv. Civil Eng., 2021, 1-19. https://doi.org/10.1155/2021/2565488.
  65. Sharif Razavian, A., Azizpour, H., Sullivan, J. and Carlsson, S. (2014). CNN features off-the-shelf: an astounding baseline for recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.
  66. Sharma, B. and Bora, P.K. (2003). "Plastic limit, liquid limit and undrained shear strength of soil-reappraisal", J. Geotech. Geoenviron. Eng., 129(8), 774-777. https://doi.org/10.1061/(ASCE)1090-0241(2003)129:8(774).
  67. Shin, H.C., Roth, H.R., Gao, M., Lu, L., Xu, Z., Nogues, I., Yao, J., Mollura, D. and Summers, R.M. (2016), "Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning", IEEE T. Med. Imaging, 35(5), 1285-1298. https://doi.org/10.1109/TMI.2016.2528162.
  68. Stefanow, D. and Dudzinski, P.A. (2021). "Soil shear strength determination methods-State of the art", Soil Tillage Res., 208, 104881. https://doi.org/10.1016/j.still.2020.104881.
  69. Thirugnanam, H. (2023), "Deep learning in landslide studies: A review", Progress in Landslide Res. Tech., 1(2), 2022, 247-255.
  70.  Tripathi, M. (2021). "Analysis of convolutional neural network based image classification techniques", J. Innov. Image Process., 3(2), 100-117. 10.36548/jiip.2021.2.003.
  71. Uncuoglu, E., Latifoglu, L. and Kaya, Z. (2023), "A hybrid approach to predict the bearing capacity of a square footing on a sand layer overlying clay", Geomech. Eng., 34(5), 561-575. https://doi.org/10.12989/gae.2023.34.5.561.
  72. Wang, Y., Fang, Z. and Hong, H. (2019), "Comparison of convolutional neural networks for landslide susceptibility mapping in Yanshan County, China", Sci. Total Environ., 666, 975-993. https://doi.org/10.1016/j.scitotenv.2019.02.263.
  73. Watabe, Y., Sassa, S., Kaneko, T. and Nakata, Y. (2017), "Mechanical characteristics of undisturbed coral gravel soils: The intergranular void ratio as a common governing parameter", Soils Found., 57(5), 760-775. https://doi.org/10.1016/j.sandf.2017.08.007.
  74. Xia, Y., Liu, C., Li, Y. and Liu, N. (2017), "A boosted decision tree approach using Bayesian hyper-parameter optimization for credit scoring", Exp. Syst. Appl., 78, 225-241. https://doi.org/10.1016/j.eswa.2017.02.017.
  75. Yoseph, H. (2022), Estmation of soil shear strength parameters from index properties using ANN the case of ADDIS ABABA.
  76.  Yoshida, Y., Kuwano, J. and Kuwano, R. (1991), "Effects of saturation on shear strength of soils", Soils Found., 31(1), 181-186. https://doi.org/10.3208/sandf1972.31.181.
  77. Zakharov, A., Shenkman, R., Ofrikhter, I. and Ponomaryov, A. (2022), "Estimation of soil properties by an artificial neural network", Mag. Civil Eng., 110(2), 11011. https://doi.org/10.34910/MCE.110.11.
  78. Zhang, W., Wu, C., Zhong, H., Li, Y. and Wang, L. (2021), "Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization", Geosci. Front., 12(1), 469-477. https://doi.org/10.1016/j.gsf.2020.03.007.
  79. Zhong, Y., Zhao, L., Liu, Z., Xu, Y. and Li, R. (2010), "Using a support vector machine method to predict the development indices of very high water cut oilfields", Petroleum Sci., 7, 379-384. https://doi.org/10.1007/s12182-010-0081-1.
  80. Zhou, J., Li, E., Wei, H., Li, C., Qiao, Q. and Armaghani, D.J. (2019), "Random forests and cubist algorithms for predicting shear strengths of rockfill materials", Appl. Sci., 9(8), 1621. https://doi.org/10.3390/app9081621.
  81. Zhu, L., Liao, Q., Wang, Z., Chen, J., Chen, Z., Bian, Q. and Zhang, Q. (2022), "Prediction of soil shear Strength parameters using combined data and different machine learning models", Appl. Sci., 12(10), 5100. https://doi.org/10.3390/app12105100.