Inteligencia Artificial aplicada a la ingeniería geotécnica: Comparación de LightGBM y XGBoost para la predicción de propiedades mecánicas del suelo
DOI:
https://doi.org/10.53673/th.v5i1.422Palabras clave:
Inteligencia artificial, Machine learning, LightGBM, XGBoost, Propiedades mecánicas del suelo, Modelos de predicción, Carreteras.Resumen
Este estudio investiga las capacidades no lineales de los modelos de predicción LightGBM y XGBoost, técnicas de machine learning dentro del campo de la Inteligencia Artificial (IA), para estimar cuatro propiedades mecánicas clave del suelo: la máxima densidad seca (MDS), el contenido óptimo de humedad (OCH), la relación de soporte de California al 100% (CBR100) y al 95% (CBR95) de la MDS. Determinar estas propiedades es fundamental en el diseño estructural de pavimentos, pero los métodos tradicionales de laboratorio suelen ser costosos y requieren tiempo. En este trabajo se utilizaron 201 registros experimentales obtenidos del Laboratorio Nº2 de Mecánica de Suelos de la Facultad de Ingeniería Civil de la UNI (FIC-UNI), incorporando como variables de entrada tres parámetros granulométricos (grava, arena y finos) y tres relacionados con los límites de consistencia (límite líquido, límite plástico e índice de plasticidad). Ambos modelos fueron entrenados utilizando optimización por grid search para ajustar sus hiperparámetros y maximizar el rendimiento predictivo. Los resultados, evaluados mediante métricas como RMSE, MAE y MAPE, indican que el modelo XGBoost supera en precisión al modelo LightGBM para la predicción de las cuatro propiedades mecánicas analizadas.
Citas
Banerjee, A., Duflo, E., Qian, N., 2020. On the road: access to transportation infrastructure and economic growth in China. J. Dev. Econ. 145 https://doi.org/ 10.1016/j.jdeveco.2020.102442.
Banerjee, A., Duflo, E., Qian, N., 2012. On the road: access to transportation infrastructure and economic growth in China. Work. Pap. 50.
Ghani, E., Goswami, A.G., Kerr, W.R., 2016. Highway to success: the impact of the Golden quadrilateral project for the location and performance of Indian manufacturing. Econ. J. 126 (591), 317–357. https://doi.org/10.1111/ecoj.12207.
Calderón, C., Servén, L., 2014. Infrastructure, Growth, and Inequality: an Overview, Policy Research Working Papers. https://doi.org/10.1596/1813-9450-7034.
Agrawal, A., Galasso, A., Oettl, A., 2017. Roads and innovation. Rev. Econ. Stat. 99 (3), 417–434. https://doi.org/10.1162/REST_a_00619.
Gibbons, S., Lyytik¨ainen, T., Sanchis-Guarner, R., 2019. New road infrastructure: the effects on firms. J. Urban Econ. 110, 35–50.
[B M. Das and N. Sivakugan, Principles of Foundation Engineering, 9th ed. USA: Cengage Learning, 2019.
[A. Özyankı, “Soil classification by using artificial neural networks,” Master’s thesis, Near East University, Nicosi, 2019.
A. Raad Al-Adhadh, H. Kadhem Sakban, and Z. Tawfiq Naeem, “Effect of Method of Soil Drying On Atterberg Limits and Soil Classification,” IOP Conf. Ser. Mater. Sci. Eng., vol. 739, no. 1, Feb. 2020, : 10.1088/1757-899X/739/1/012044.
B. Zhou and N. Lu, “Correlation between Atterberg Limits and Soil Adsorptive Water,” J. Geotech. Geoenvironmental Eng., vol. 147, no. 2, Feb. 2021, doi: 10.1061/(ASCE)GT.1943- 5606.0002463.
V. Y. Katte, S. M. Mfoyet, B. Manefouet, A. S. L. Wouatong, and L. A. Bezeng, “Correlation of California Bearing Ratio (CBR) Value with Soil Properties of Road Subgrade Soil,” Geotech. Geol. Eng., vol. 37, no. 1, pp. 217–234, Jan. 2019, doi: 10.1007/s10706-018-0604-x.
T. Li, L. Kong, and B. Liu, “The California Bearing Ratio and Pore Structure Characteristics of Weakly Expansive Soil in Frozen Areas,” Appl. Sci., vol. 10, no. 21, pp. 1–22, Oct. 2020, doi: 10.3390/app10217576.
X. Ji et al., “A Prediction Method for the California Bearing Ratio of Soil-Rock Mixture Based on the Discrete Element Method and CT Scanning,” Adv. Civ. Eng., vol. 2020, pp. 1–12, Sep. 2020, doi: 10.1155/2020/9794756.
E. Mina, R. I. Kusuma, and N. Ulfah, “Utilization of steel slag and fly ash in soil stabilization and their effect to california bearing ratio (CBR) value. (Case study: Kp. Kadusentar road Medong village Mekarjaya Subdistrict Pandeglang District),” in IOP Conference Series: Materials Science and Engineering, Dec. 2019, vol. 673, no. 1, doi: 10.1088/1757-899X/673/1/012034.
E. A. Sandoval Vallejo y W. A. Rivera Mena, “Correlación del CBR con la resistencia a la compresión inconfnada”, Cienc. e Ing. Neogranadina, vol. 29, no. 1, pp. 135–152, Aug. 2019, doi: 10.18359/rcin.3478.
Universidad Nacional de Ingeniería, “Tarifa de ensayos de laboratorio”, Lima, 2019. [En línea]. Disponible en: http://www.lms.uni.edu.pe/labsuelos/tarifa/tarifa general 2019.pdf.
Bogado, G. O., Pintos, N. A., Reinert, H. O., & Bressan, D. A. (2017). Correlación entre parámetros de compactación y propiedades geotécnicas en suelos tropicales misioneros.
Ministerio de Transportes y Comunicaciones, “Manual de carreteras–Suelos, Geología, Geotécnia y Pavimentos. Sección: Suelos y Pavimentos”. Perú, 2014.
E. Özgan, S. Serin, and İ. Vural, “Multi-faceted investigation and modeling of compaction parameters for road construction,” J. Terramechanics, vol. 60, pp. 33–42, Aug. 2015, doi: 10.1016/j.jterra. 2015.02.005.
Vafoeva, O. (2020, July). Hydromechanical Method of Soil Compaction. In IOP Conference Series: Materials Science and Engineering (Vol. 883, No. 1, p. 012061). IOP Publishing.
J. C. Chirinos Quispe, “Efecto de la energía de compactación en la densidad seca máxima y contenido óptimo de humedad del suelo granular de la cantera El Gavilán, 2015”, Tesis de licenciatura, Universidad Privada del Norte, 2016.
E. Millan-Romero and C. Millan-Paramo, “Relationship of water parameters with the optimum moisture content in clay soils,” ARPN J. Eng. Appl. Sci., vol. 15, no. 15, pp. 1666–1671, 2020.
I. Ekanayake, D. Meddage, U. Rathnayake, A novel approach to explain the black-box nature of machine learning in compressive strength predictions of concrete using Shapley additive explanations (SHAP), Case Studies, Constr. Mater. (2022) e01059.
N. Kardani, A. Bardhan, D. Kim, P. Samui, A. Zhou, Modelling the energy performance of residential buildings using advanced computational frameworks based on RVM GMDH, ANFIS-BBO and ANFIS-IPSO, J. Build. Eng. 35 (2021), 102105.
M.N.A. Raja, S.K. Shukla, Predicting the settlement of geosynthetic-reinforced soil foundations using evolutionary artificial intelligence technique, Geotextiles and Geomembranes 49 (5) (2021) 1280–1293.
B. Hilloulin, V.Q. Tran, Using machine learning techniques for predicting autogenous shrinkage of concrete incorporating superabsorbent polymers and supplementary cementitious materials, J. Build. Eng. (2022), 104086.
H. Adel, M.I. Ghazaan, A.H. Korayem, Machine learning applications for developing sustainable construction materials, Artificial Intelligence and Data Science in Environmental Sensing, Elsevier (2022) 179–210.
J. de Brito, C. Thomas, C. Medina, F. Agrela, Waste and byproducts in cement-based materials: innovative sustainable materials for a circular economy, Woodhead Publishing, 2021.
F. Pacheco-Torgal, L.F. Cabeza, J. Labrincha, A.G. De Magalhaes, Eco-efficient construction and building materials: life cycle assessment (LCA), eco-labelling and case studies, woodhead Publishing, 2014.
C. Rajakumar and G. Reddy Babu, “Experimental study and neural network modelling of expansive sub grade stabilized with industrial waste byproducts and geogrid,” Mater. Today Proc., Aug. 2020, doi: 10.1016/j.matpr.2020.06.578
T. Fikret Kurnaz and Y. Kaya, “Prediction of the California bearing ratio (CBR) of compacted soils by using GMDH-type neural network,” Eur. Phys. J. Plus, vol. 134, no. 7, Jul. 2019, doi: 10.1140/ epjp/i2019-12692-0.
M. Suthar and P. Aggarwal, “Predicting CBR Value of Stabilized Pond Ash with Lime and Lime Sludge Using ANN and MR Models,” Int. J. Geosynth. Gr. Eng., vol. 4, no. 1, p. 6, Mar. 2018, doi: 10.1007/s40891-017-0125-3.
F. J. Valderrama Purizaca, D. A. Chávez Barturen, S. P. Muñoz Pérez, V. A. Tuesta-Monteza, and H. I. Mejía-Cabrera, “Importance of artificial neural networks in civil engineering: a systematic review of the literature,” ITECKNE, vol. 18, no. 1, 2020, doi: 10.15332/iteckne.v18i1.2542.
V.Q. Tran, Machine learning approach for investigating chloride diffusion coefficient of concrete containing supplementary cementitious materials, Constr. Build. Mater. 328 (2022), 127103.
J. Liang, Y. Bu, K. Tan, J. Pan, Z. Yi, X. Kong, Z. Fan, Estimation of stellar atmospheric parameters with light gradient boosting machine algorithm and principal component analysis, The Astronomical Journal 163 (4) (2022) 153.
X. Hao, Z. Zhang, Q. Xu, G. Huang, K. Wang, Prediction of f-CaO content in cement clinker: A novel prediction method based on LightGBM and Bayesian optimization, Chemometrics and Intelligent Laboratory Systems 220 (2022), 104461.
C. Zhang, X. Lei, L. Liu, Predicting metabolite-disease associations based on LightGBM model, Frontiers in Genetics 12 (2021) 446.
P.j. Chun, S. Izumi, T. Yamane, Automatic detection method of cracks from concrete surface imagery using two-step light gradient boosting machine, Computer-Aided Civil and Infrastructure Engineering 36(1) (2021) 61-72.
T. Chen, T. He, M. Benesty, V. Khotilovich, Y. Tang, H. Cho, Xgboost: extreme gradient boosting, R package version 0.4-2 1 (4) (2015) 1–4.
J. Liu, S. Zhang, H. Fan, A two-stage hybrid credit risk prediction model based on XGBoost and graph-based deep neural network, Expert Systems with Applications 195 (2022), 116624.
E. Al Daoud, Comparison between XGBoost, LightGBM and CatBoost using a home credit dataset, Int. J. Computer Inform. Eng. 13 (1) (2019) 6–10.
H. Nguyen, X.-N. Bui, H.-B. Bui, D.T. Cuong, Developing an XGBoost model to predict blast-induced peak particle velocity in an open-pit mine: a case study, Acta Geophysica 67 (2) (2019) 477–490.
J. Duan, P.G. Asteris, H. Nguyen, X.-N. Bui, H. Moayedi, A novel artificial intelligence technique to predict compressive strength of recycled aggregate concrete using ICA-XGBoost model, Engineering with Computers 37 (4) (2021) 3329–3346.
H. Imran, M. Ibrahim, S. Al-Shoukry, F. Rustam, I. Ashraf, Latest concrete materials dataset and ensemble prediction model for concrete compressive strength containing RCA and GGBFS materials, Constr. Build. Mater. 325 (2022), 126525
J. Liu, S. Zhang, H. Fan, A two-stage hybrid credit risk prediction model based on XGBoost and graph-based deep neural network, Expert Systems with Applications 195 (2022), 116624.
N.-H. Nguyen, J. Abell´an-García, S. Lee, E. Garcia-Castano, T.P. Vo, Efficient estimating compressive strength of ultra-high performance concrete using XGBoost model, J. Build. Eng. (2022), 104302.
M.N. Amin, M. Iqbal, K. Khan, M.G. Qadir, F.I. Shalabi, A. Jamal, Ensemble treebased approach towards flexural strength prediction of FRP reinforced concrete beams, Polymers 14 (7) (2022) 1303.
A. Bardhan, P. Samui, K. Ghosh, A.H. Gandomi, S. Bhattacharyya, ELM-based adaptive neuro swarm intelligence techniques for predicting the California bearing ratio of soils in soaked conditions, Applied Soft Computing 110 (2021), 107595.
A. Bardhan, C. Gokceoglu, A. Burman, P. Samui, P.G. Asteris, Efficient computational techniques for predicting the California bearing ratio of soil in soaked conditions, Engineering Geology 291 (2021), 106239.
L. Sun, M. Koopialipoor, D.J. Armaghani, R. Tarinejad, M. Tahir, Applying a metaheuristic algorithm to predict and optimize compressive strength of concrete samples, Engineering with Computers (2019) 1–13.
W. Liang, S. Luo, G. Zhao, H. Wu, Predicting hard rock pillar stability using GBDT, XGBoost, and LightGBM algorithms, Mathematics 8 (5) (2020) 765.
H. Zeng, C. Yang, H. Zhang, Z. Wu, J. Zhang, G. Dai, F. Babiloni, W. Kong, A lightGBM-based EEG analysis method for driver mental states classification, Computational Intelligence and Neuroscience 2019 (2019).
H. Kodaz, S. ¨Ozs¸en, A. Arslan, S. Günes¸, Medical application of information gain based artificial immune recognition system (AIRS): Diagnosis of thyroid disease, Expert Systems with Applications 36 (2) (2009) 3086–3092.
T. Chen, C. Guestrin, Xgboost: A scalable tree boosting system, in: Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 2016, pp. 785–794.
J. Fan, X. Wang, L. Wu, H. Zhou, F. Zhang, X. Yu, X. Lu, Y. Xiang, Comparison of Support Vector Machine and Extreme Gradient Boosting for predicting daily global solar radiation using temperature and precipitation in humid subtropical climates: A case study in China, Energy Convers. Manag. 164 (2018) 102–111.
I. Ullah K. Liu T. Yamamoto R.E. Al Mamlook A. Jamal A comparative performance of machine learning algorithm to predict electric vehicles energy consumption: A path towards sustainability Energy & Environment 2021 0958305X211044998.
M.I. Khan, M.H. Sutanto, K. Khan, M. Iqbal, M.B. Napiah, S.E. Zoorob, J.J. Klemeˇs, A. Bokhari, W. Rafiq, Effective use of recycled waste PET in cementitious grouts for
M. Iqbal, Q. Zhao, D. Zhang, F.E. Jalal, A. Jamal, Evaluation of tensile strength degradation of GFRP rebars in harsh alkaline conditions using non-linear genetic-based models, Materials and Structures 54 (5) (2021) 1–20.
M. Iqbal, D. Zhang, F.E. Jalal, M., Faisal Javed, Computational AI prediction models for residual tensile strength of GFRP bars aged in the alkaline concrete environment, Ocean Engineering 232 (2021), 109134. [70] M.I. Khan, M.H. Sutanto, K. Khan, M. Iqbal, M.B. Napiah
F.E. Jalal, Y. Xu, M. Iqbal, B. Jamhiri, M.F. Javed, Predicting the compaction characteristics of expansive soils using two genetic programming-based algorithms, Transportation Geotechnics 30 (2021), 100608.
Kodaz, H., Özşen, S., Arslan, A., & Güneş, S. (2009). Medical application of information gain based artificial immune recognition system (AIRS): Diagnosis of thyroid disease. Expert Systems with Applications, 36(2), 3086-3092.
Carrillo, A. (2016). Comportamiento del suelo tropical peruano. Universidad Ricardo Palma. Lima–Perú. Pág, 1-35.
Publicado
Cómo citar
Número
Sección
Licencia
Esta obra está bajo una licencia internacional Creative Commons Atribución 4.0.
El autor permite el uso y difusión de los artículos a los lectores y a la revista, por medio de la licencia Creative Commons CC-BY-4.0 (Reconocimiento 4.0). Esta licencia permite que cualquiera distribuya, remezcle, adapte y rediseñe a partir de un trabajo, incluso con fines lucrativos, siempre que se mencione la autoría de la creación original.
Ruta CC-BY-4.0: https://creativecommons.org/licenses/by/4.0/
