Geomechanics and Engineering

  • 제39권2호
  • /
  • Pages.129-142
  • /
  • 2024
  • /
  • 2005-307X(pISSN)
  • /
  • 2092-6219(eISSN)
테크노프레스 (Techno-Press)
권호 표지
DOI QR코드

DOI QR Code

Using multivariate regression and multilayer perceptron networks to predict soil shear strength parameters

  • Ahmed Cemiloglu (School of Information Engineering, Yancheng Teachers University)
  • 투고 : 2024.04.10
  • 심사 : 2024.09.20
  • 발행 : 2024.10.25
⟨ 이전 논문 다음 논문 ⟩

초록

The most significant soil parameters that are utilized in geotechnical engineering projects' design and implementations are soil strength parameters including friction (ϕ), cohesion (c), and uniaxial compressive strength (UCS). Understanding soil shear strength parameters can be guaranteed the design success and stability of structures. In this regard, professionals always looking for ways to get more accurate estimations. The presented study attempted to investigate soil shear strength parameters by using multivariate regression and multilayer perceptron predictive models which were implemented on 100 specimens' data collected from the Tabriz region (NW of Iran). The uniaxial (UCS), liquid limit (LL), plasticity index (PI), density (γ), percentage of fine-grains (pass #200), and sand (pass #4) which are used as input parameters of analysis and shear strength parameters predictions. A confusion matrix was used to validate the testing and training data which is controlled by the coefficient of determination (R2), mean absolute (MAE), mean squared (MSE), and root mean square (RMSE) errors. The results of this study indicated that MLP is able to predict the soil shear strength parameters with an accuracy of about 93.00% and precision of about 93.5%. In the meantime, the estimated error rate is MAE = 2.0231, MSE = 2.0131, and RMSE = 2.2030. Additionally, R2 is evaluated for predicted and measured values correlation for friction angle, cohesion, and UCS are 0.914, 0.975, and 0.964 in the training dataset which is considerable.

키워드

과제정보

The author extends his gratitude to the Editorial Team of GAE Journal for their efforts and considerations. Also, this research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. The authors have not disclosed any competing interests.

참고문헌

  1. Abdalla, J.A., Attom, M.F. and Hawileh, R. (2015), "Prediction of minimum factor of safety against slope failure in clayey soils using artificial neural network", Environ. Earth Sci., 73, 5463-5477. https://doi.org/10.1007/s12665-014-3800-x. 
  2. Abramson, L.W., Lee, T.S., Sharma, S. and Boyce, G.M. (2001), Slope stability and stabilization methods, John Wiley & Sons, New York, NY, USA. 
  3. Abunama, T., Othman, F., Ansari, M. and El-Shafie, A. (2019), "Leachate generation rate modeling using artificial intelligence algorithms aided by input optimization method for an MSW landfill", Environ. Sci. Poll. Res., 8(26), 3368-3381. https://doi.org/10.1007/s11356-018-3749-5. 
  4. ASTM D2166 (2006), Standard Test Method for Unconfined Compressive Strength of Cohesive Soil, ASTM International, West Conshohocken, PA, USA. 
  5. ASTM D4318 (2017), Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils, ASTM International, West Conshohocken, PA, USA. 
  6. ASTM D6913 (2017), Standard Test Methods for Particle-Size Distribution (Gradation) of Soils Using Sieve Analysis, ASTM International, West Conshohocken, PA, USA. 
  7. ASTM D7263 (2021), Standard Test Methods for Laboratory Determination of Density and Unit Weight of Soil Specimens, ASTM International, West Conshohocken, PA, USA. 
  8. Azadi, A., Irani, A.E., Azarafza, M., Hajialilue Bonab, M., Sarand, F.B. and Derakhshani, R. (2022), "Coupled numerical and analytical stability analysis charts for an earth-fill dam under rapid drawdown conditions", Appl. Sci., 12(9), 4550. https://doi.org/10.3390/app12094550. 
  9. Azarafza, M. and Mokhtari, M.H. (2013), "Evaluation drought effect on the Urmia lake salinity change by using remote sensing techniques", J. Arid. Biomed, 3(2), 1-14. 
  10. Azarafza, M., Ghazifard, A., Akgun, H. and Asghari-Kaljahi, E. (2019), "Geotechnical characteristics and empirical geoengineering relations of the South Pars Zone marls, Iran", Geomech. Eng., 19(5), 393-405. https://doi.org/10.12989/gae.2019.19.5.393. 
  11. Azarafza, M., Hajialilue Bonab, M. and Derakhshani, R. (2022a), "A novel empirical classification method for weak rock slope stability analysis", Sci. Rep., 12(1), 14744. https://doi.org/10.1038/s41598-022-19246-w. 
  12. Azarafza, M., Hajialilue Bonab, M. and Derakhshani, R. (2022b), "A deep learning method for the prediction of the index mechanical properties and strength parameters of marlstone", Materials, 15(19), 6899. https://doi.org/10.3390/ma15196899. 
  13. Azarafza, M., Nikoobakht, S., Rahnamarad, J., Asasi, F. and Derakhshani, R. (2020c), "An empirical method for slope mass rating-Qslope correlation for Isfahan province, Iran", MethodsX, 7, 101069. https://doi.org/10.1016/j.mex.2020.101069. 
  14. Bahnsen, A.C. and Gonzalez, A.M. (2011), "Evolutionary algorithms for selecting the architecture of a MLP neural network: A credit scoring case", Proceedings of the 2011 IEEE 11th International Conference on Data Mining Workshops, Vancouver, Canada, December. 
  15. Baykasoglu, A., Gullu, H., Canakci, H. and Ozbakir, L. (2008), "Prediction of compressive and tensile strength of limestone via genetic programming", Exp. Syst. Appl., 35(1-2), 111-123. https://doi.org/10.1016/j.eswa.2007.06.006. 
  16. Canakci, H., Baykasoglu, A. and Gullu, H., (2009), "Prediction of compressive and tensile strength of Gaziantep basalts via neural networks and gene expression programming", Neural Comput. Appl., 18, 1031-1041. https://doi.org/10.1007/s00521-008-0208-0. 
  17. Cemiloglu, A., Zhu, L., Arslan, S., Xu, J., Yuan, X., Azarafza, M. and Derakhshani, R. (2023), "Support Vector Machine (SVM) Application for Uniaxial Compression Strength (UCS) Prediction: A case study for Maragheh Limestone", Appl. Sci., 13(4), 2217. https://doi.org/10.3390/app13042217. 
  18. Chicco, D. and Jurman, G. (2020), "The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation", BMC Genomics, 21, 1-13. https://doi.org/10.1186/s12864-019-6413-7. 
  19. Chicco, D., Warrens, M.J. and Jurman, G. (2021), "The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation", Peerj Comput. Sci., 7, e623. https://doi.org/10.7717/peerj-cs.623. 
  20. Das, S.K. and Basudhar, P.K. (2008), "Prediction of residual friction angle of clays using artificial neural network", Eng. Geol., 100, 142-145. https://doi.org/10.1016/j.enggeo.2008.03.001. 
  21. Day, R.W. (2012), Geotechnical engineer's portable handbook, McGraw-Hill Education, Berkshire, UK. 
  22. Dehghan, S., Sattari, G., Chelgani, S.C. and Aliabadi, M.A. (2010), "Prediction of uniaxial compressive strength and modulus of elasticity for Travertine samples using regression and artificial neural networks", Min. Sci. Technol. (China), 20(1), 41-46. https://doi.org/10.1016/S1674-5264(09)60158-7. 
  23. Diamantopoulou, M.J. (2005), "Artificial neural networks as an alternative tool in pine bark volume estimation", Comput. Electron. Agr., 48(3), 235-244. https://doi.org/10.1016/j.compag.2005.04.002. 
  24. Dinarvand, R. and Ardakan, A. (2022), "Shear behavior of geotextile-encased gravel columns in silty sand-Experimental and SVM modeling", Geomech. Eng., 28(5), 505-520. https://doi.org/10.12989/gae.2022.28.5.505. 
  25. Fang, H.Y. (2013), Foundation engineering handbook, Springer Science & Business Media, Berlin/Heidelberg, Germany.
  26. Germaine, J.T. and Germaine, A.V. (2009), Geotechnical laboratory measurements for engineers. John Wiley & Sons, New York, NY, USA. 
  27. Gunaydin, O., Gokoglu, A. and Fener, M. (2010), "Prediction of artificial soil's unconfined compression strength test using statistical analyses and artificial neural networks", Adv. Eng. Softw., 41(9), 1115-1123. https://doi.org/10.1016/j.advengsoft.2010.06.008. 
  28. Hahn, T., Marquand, A.F., Plichta, M.M., Ehlis, A.C., Schecklmann, M.W., Dresler, T., Jarczok, T.A., Eirich, E., Leonhard, C., Reif, A., Lesch, K.P., Brammer, M.J., Miranda, J.M. and Fallgatter, A.J. (2012), "A novel approach to probabilistic biomarker-based classification using functional near-infrared spectroscopy", Human Brain Mapping, 34(5), 1102-1114. https://doi.org/10.1002/hbm.21497. 
  29. Hearty, J. (2016), Advanced Machine Learning with Python, Packt Publishing, Birmingham, UK. 
  30. Hodson, T.O. (2022), "Root-mean-square error (RMSE) or mean absolute error (MAE): When to use them or not", Geosci. Model Devel., 15(14), 5481-5487. https://doi.org/10.5194/gmd-15-5481-2022. 
  31. Hooshmand, A., Aminfar, M.H., Asghari, E. and Ahmadi, H. (2012), "Mechanical and physical characterization of Tabriz marls, Iran", Geotech. Geol. Eng., 30, 219-232. https://doi.org/10.1007/s10706-011-9464-3. 
  32. Ibnu Choldun R.M., Santoso, J. and Surendro, K. (2020), "Determining the number of hidden layers in neural network by using principal component analysis", Proceedings of the 2019 Intelligent Systems Conference (IntelliSys). https://doi.org/10.1007/978-3-030-29513-4_36. 
  33. Kahraman, S., Gunaydin, O., Alber, M. and Fener, M. (2009), "Evaluating the strength and deformability properties of Misis fault breccia using artificial neural networks", Exp. Syst. Appl., 36(3), 6874-6878. https://doi.org/10.1016/j.eswa.2008.08.002. 
  34. Karumanchi, M. and Nerella, R. (2022), "Shear strength parameters from digital tri-axial test and soils stabilization with extracted nanosilica", Nanotech. Environ. Eng., 7(1), 307-318. https://doi.org/10.1007/s41204-022-00238-0. 
  35. Kayadelen, C., Gunaydin, O., Fener, M., Demir, A. and Ozvan, A. (2009), "Modeling of the angle of shearing resistance of soils using soft computing systems", Exp. Syst. Appl., 36, 11814-11826. https://doi.org/10.1016/j.eswa.2009.04.008. 
  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. Kiranyaz, S., Avci, O., Abdeljaber, O., Ince, T., Gabbouj, M. and Inman, D.J. (2021), "1D convolutional neural networks and applications: A survey", Mech. Syst. Signal Pr., 151, 107398. https://doi.org/10.1016/j.ymssp.2020.107398. 
  38. Liew, S.S., Khalil-Hani, M. and Bakhteri, R. (2016), "Bounded activation functions for enhanced training stability of deep neural networks on visual pattern recognition problems", Neurocomput., 216, 718-734. https://doi.org/10.1016/j.neucom.2016.08.037. 
  39. Mao, Y., Azarafza, M., Bonab, M.H., Bascompta Massanes, M. and Nanehkaran, Y.A. (2023), "Empirical correlation for in-situ deformation modulus of sedimentary rock slope mass and support system recommendation using the Qslope method", Geomech. Eng., 35(5), 539-554. https://doi.org/10.12989/gae.2023.35.5.539. 
  40. Mao, Y., Chen, L., Nanehkaran, Y.A., Azarafza, M. and Derakhshani, R. (2023), "Fuzzy-based intelligent model for rapid rock slope stability analysis using Qslope", Water, 15(16), 2949. https://doi.org/10.3390/w15162949. 
  41. Mertler, C.A., Vannatta, R.A. and LaVenia, K.N. (2021), Advanced and multivariate statistical methods: Practical application and interpretation, Routledge, London, England, UK. 
  42. Moayedi, H., Gor, M., Khari, M., Foong, L.K., Bahiraei, M. and Bui, D.T. (2020), "Hybridizing four wise neural-metaheuristic paradigms in predicting soil shear strength", Measurement, 156, 107576. https://doi.org/10.1016/j.measurement.2020.107576. 
  43. Mohammadi, M., Aghda S.M.F., Talkhablou, M. and Cheshomi 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. 
  44. Mohammadi, S.D., Naseri, F. and Alipoor, S. (2014), "Development of artificial neural networks and multiple regression models for the NATM tunnelling-induced settlement in Niayesh subway tunnel, Tehran", Bull. Eng. Geol. Environ., 74, 827-884. https://doi.org/10.1007/s10064-014-0660-2. 
  45. Mokkadem, A., Pelletier, M. and Thiam, B. (2008), "Large and moderate deviations principles for kernel estimators of the multivariate regression", Math. Method. Stat., 17, 146-172. https://doi.org/10.3103/S1066530708020051. 
  46. Murthy, V.N.S. (2002), Geotechnical engineering: principles and practices of soil mechanics and foundation engineering. CRC press, Boca Raton, FL, USA. 
  47. Nanehkaran, Y.A., Mao, Y., Azarafza, M., Kockar, M.K. and Zhu, H.H. (2021), "Fuzzy-based multiple decision method for landslide susceptibility and hazard assessment: A case study of Tabriz, Iran", Geomech. Eng., 24(5), 407-418. https://doi.org/10.12989/gae.2021.24.5.407. 
  48. Nanehkaran, Y.A., Pusatli, T., Chengyong, J., Chen, J., Cemiloglu, A., Azarafza, M. and Derakhshani, R. (2022), "Application of machine learning techniques for the estimation of the safety factor in slope stability analysis", Water, 14(22), 3743. https://doi.org/10.3390/w14223743. 
  49. Nguyen, Q.H., Ly, H.B., Ho, L.S., Al-Ansari, N., Le, H.V. and Tran, V.Q. (2021), "Influence of data splitting on performance of machine learning models in prediction of shear strength of soil", Math. Probl. Eng., 2021, 4832864. https://doi.org/10.1155/2021/4832864. 
  50. Park, Y.S. and Lek, S. (2016), "Chapter 7 - Artificial Neural Networks: Multilayer Perceptron for Ecological Modeling Author links open overlay panel", Develop. Environ. Model., 28, 123-140. https://doi.org/10.1016/B978-0-444-63623-2.00007-4. 
  51. 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. 
  52. Pham, B.T., Nguyen, M.D., Nguyen-Thoi, T., Ho, L.S., Koopialipoor, M., Quoc, N.K. and Van Le, H. (2021), "A novel approach for classification of soils based on laboratory tests using Adaboost, Tree and ANN modeling", Transport. Geotech., 27, 100508. https://doi.org/10.1016/j.trgeo.2020.100508. 
  53. Pitilakis, K., Raptakis, D., Lontzetidis, K., Tika-Vassilikou, T. and Jongmans, D. (1999), "Geotechnical and geophysical description of EURO-SEISTEST, using field, laboratory tests and moderate strong motion recordings", J. Earthq. Eng., 3(3), 381-409. https://doi.org/10.1080/13632469909350352. 
  54. Rights, J.D. and Sterba, S.K. (2019), "Quantifying explained variance in multilevel models: An integrative framework for defining R-squared measures", Psychol. Methods, 24(3), 309. https://doi.org/10.1037/met0000184. 
  55. Sedgwick, P. (2012), "Pearson's correlation coefficient". Bmj, 2012, 345. https://doi.org/10.1136/bmj.e4483. 
  56. Sharma, S., Ahmed, S., Naseem, M., Alnumay, W.S., Singh, S. and Cho, G.H. (2021), "A survey on applications of artificial intelligence for pre-parametric project cost and soil shear-strength estimation in construction and geotechnical engineering", Sensors, 21(2), 463. https://doi.org/10.3390/s21020463. 
  57. Sharma, S., Sharma, S. and Athaiya, A. (2017), "Activation functions in neural networks", Towards Data Sci., 6(12), 310-316. 
  58. Soltanian, A., Zad A., Yazdi, M. and Tohidi, A. (2024), "An experimental investigation on dispersion and geotechnical properties of dispersive clay soil stabilized with Metakaolin and Zeolite", Geomech. Eng., 36(6), 589-599. https://doi.org/10.12989/gae.2024.36.6.589. 
  59. Szandala, T. (2021), "Review and comparison of commonly used activation functions for deep neural networks", Bio-inspired Neurocomput., 203-224. https://doi.org/10.48550/arXiv.2010.09458. 
  60. Torabi-Kaveh, M. and Sarshari, B. (2019), "Predicting Convergence Rate of Namaklan Twin Tunnels Using Machine Learning Methods", Arab. J. Sci. Eng., 45, 3761-3780. https://doi.org/10.1007/s13369-019-04239-1. 
  61. Valizadeh, A. (2021), "A way to reduce the time consumption effect of for-loops for training neural networks: Optimized propagation", Research Square, PPR379229. https://doi.org/10.21203/rs.3.rs-776504/v3. 
  62. Venkatesh, K. and Bind, Y.K. (2020), "ANN and neuro-fuzzy modeling for shear strength characterization of soils", Proc. Nat. Academy Sci., India Sec. A: Phys. Sci., 92, 243-249. https://doi.org/10.1007/s40010-020-00709-6. 
  63. Weerakody, P.B., Wong, K.W., Wang, G. and Ela, W. (2021), "A review of irregular time series data handling with gated recurrent neural networks", Neurocomputing, 441, 161-178. https://doi.org/10.1016/j.neucom.2021.02.046. 
  64. Yilmaz, I. and Kaynar, O. (2011), "Multiple regression, ANN (RBF, MLP) and ANFIS models for prediction of swell potential of clayey soils", Exp. Syst. Appl., 38(5), 5958-5966. https://doi.org/10.1016/j.eswa.2010.11.027. 
  65. 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. 
  66. Zhang, H., Nguyen, H., Bui, X.N., Pradhan, B., Asteris, P.G., Costache, R. and Aryal, J. (2022), "A generalized artificial intelligence model for estimating the friction angle of clays in evaluating slope stability using a deep neural network and Harris Hawks optimization algorithm", Eng. Comput., 38, 3901-3914. https://doi.org/10.1007/s00366-020-01272-9. 
  67. Zhang, X., Nguyen, H., Bui, X.N., Le, H.A., Nguyen-Thoi, T., Moayedi, H. and Mahesh, V. (2020). "Evaluating and Predicting the Stability of Roadways in Tunnelling and Underground Space Using Artificial Neural Network-Based Particle Swarm Optimization", Tunn. Undergr. Sp. Tech., 103, 103517. https://doi.org/10.1016/j.tust.2020.103517. 
  68. Zhu, W., Zeng, N. and Wang, N. (2010), "Sensitivity, specificity, accuracy, associated confidence interval and ROC analysis with practical SAS implementations", NESUG Proceedings: Healthcare Life Sci., 19, 67.