پیشبینی پتانسیل روانگرایی با رویکرد محاسبات نرم (CNN-MVO) مطالعه موردی خاکهای ماسهای شمال ایران
محورهای موضوعی : آنالیز سازه - زلزله
شیما آقاکثیری
1
,
قدرت الله محمدی
2
,
امیر تابان
3
,
Mohammad Emami korandeh
4
1 - گروه مهندسی عمران، واحد تهرانجنوب، دانشگاه آزاد اسلامی، تهران، ایران
2 - Assistant Professor; Department of Civil Engineering, South Tehran Branch, Islamic Azad University, Tehran, Iran
3 - Assistant Professor, Department of Civil Engineering, Khorramabad Branch, Islamic Azad University, Tehran, Iran
4 - Academic staff of the Civil Engineering Department and member of the Council of the Civil Engineering Department
کلید واژه: روانگرایی, یادگیری ماشین عمیق, محاسبات نرم, الگوریتم بهینهسازی (MVO), شبکه عصبی CNN.,
چکیده مقاله :
روانگرایی در خاکهای ماسهای یکی از مخاطرات مهم ژئوتکنیکی در هنگام زلزله است که میتواند منجر به خسارات جدی سازهای و ژئوتکنیکی گردد. هدف این پژوهش، توسعه یک مدل محاسبات نرم برای پیشبینی پتانسیل روانگرایی با استفاده از دادههای ژئوتکنیکی بهدستآمده از پروژههای شمال ایران است. برای این منظور، مجموعهای شامل ۱۵۰ گمانه صحرایی و دادههای حاصل از آزمایشهای SPT و پارامترهای ژئوتکنیکی شامل نوع خاک، سطح آب زیرزمینی، درصد ریزدانه، عدد نفوذ استاندارد و عمق مورداستفاده قرار گرفت. روش پیشنهادی ترکیبی از شبکه عصبی کانولوشن (CNN) و الگوریتم چندجهانی (MVO) بوده است که در محیط MATLAB پیادهسازی شد. نتایج نشان داد ضریب رگرسیون مدل توسعهیافته حدود 9/0 و شاخص خطای میانگین مربعات (MSE) کمتر از 5/0 است. مقایسه نتایج با روش تجربی مبتنی بر SPT نشان داد که مدل CNN-MVO دقت بالاتری در پیشبینی وقوع روانگرایی و نشست ناشی از آن دارد؛ بنابراین، روش ارائهشده میتواند بهعنوان ابزاری کارآمد در ارزیابی خطر روانگرایی و تصمیمگیریهای مهندسی ژئوتکنیک در مناطق مستعد زلزله مورداستفاده قرار گیرد.
Liquefaction in sandy soils is one of the most critical geotechnical hazards during earthquakes, often resulting in severe structural and infrastructural damages. The present study aims to develop a soft computing model for predicting liquefaction potential using geotechnical data obtained from case studies in northern Iran. A dataset of 150 boreholes including Standard Penetration Test (SPT) results and key soil parameters such as soil type, groundwater level, fines content, SPT number, and depth was employed. The proposed approach integrates a Convolutional Neural Network (CNN) with the Multi-Verse Optimizer (MVO) algorithm, implemented in MATLAB. The results indicate a regression coefficient (R) of approximately 0.90 and a mean squared error (MSE) lower than 0.5. Comparison with conventional SPT-based empirical methods confirmed the superior accuracy of the CNN-MVO model in predicting liquefaction occurrence and associated settlements. Therefore, the proposed methodology provides a reliable tool for liquefaction hazard assessment and geotechnical decision-making in earthquake-prone regions.
[1]. Luque, R. and J.D. Bray, Dynamic soil-structure interaction analyses of two important structures affected by liquefaction during the Canterbury earthquake sequence. Soil Dynamics and Earthquake Engineering, 2020. 133: p. 106026.
https://doi.org/10.1016/j.soildyn.2019.106026
[2]. Zhou, J., et al., Feasibility of stochastic gradient boosting approach for evaluating seismic liquefaction potential based on SPT and CPT case histories. Journal of Performance of Constructed Facilities, 2019. 33(3): p. 04019024.
https://doi.org/10.1061/(ASCE)CF.1943-5509.0001292
[3]. Kokusho, T., Innovative earthquake soil dynamics. 2017: CRC Press.
[4]. Ahmad, M., et al., A hybrid approach for evaluating CPT-based seismic soil liquefaction potential using Bayesian belief networks. J. Cent. South Univ, 2020. 27: p. 500-516
https://doi.org/10.1007/s11771-020-4312-3
[5]. Zhou, J., et al. Classification of seismic-liquefaction potential using Friedman’s stochastic gradient boosting based on the cone penetration test data. in Transportation and Geotechniques: Materials, Sustainability and Climate: Proceedings of the 5th GeoChina International Conference 2018–Civil Infrastructures Confronting Severe Weathers and Climate Changes: From Failure to Sustainability, held on July 23 to 25, 2018 in HangZhou, China. 2019. Springer.
https://doi.org/10.1007/978-3-319-95768-5_7
[6]. Nakao, K., et al., Numerical simulation of the liquefaction phenomenon by MPSM-DEM coupled CAES. Sustainability, 2022. 14(12): p. 7517 . https://doi.org/10.3390/su14127517
[7]. Cong, Y., et al., Machine learning predictive analysis of liquefaction resistance for sandy soils enhanced by chemical injection. Machine Learning and Knowledge Extraction, 2024. 6(1): p. 402-419.
https://doi.org/10.3390/make6010020
[8]. Cavus, U.S., et al., Fuzzy logic based assessment of seismic soil liquefaction potential and its application to foundations of bridge piers. Journal of Intelligent & Fuzzy Systems, 2019. 36(6): p. 6001-6011. https://doi.org/10.3233/JIFS-181795
[9]. Park, S.-S., et al., A simple and sustainable prediction method of liquefaction-induced settlement at Pohang using an artificial neural network. Sustainability, 2020. 12(10): p. 4001.
[10]. Tang, X.-W., J.-L. Hu, and J.-N. Qiu, Identifying significant influence factors of seismic soil liquefaction and analyzing their structural relationship. KSCE Journal of Civil Engineering, 2016. 20: p. 2655-2663. https://doi.org/10.1007/s12205-016-0339-2
[11]. Cetin, K. and H. Bilge, Recent advances in seismic soil liquefaction engineering. The Evolution of Geotech-25 Years of Innovation, 2021: p. 18-42.
[12]. Fang, Y., I. Jairi, and N. Pirhadi, Neural transfer learning for soil liquefaction tests. Computers & Geosciences, 2023. 171: p. 105282
https://doi.org/10.1016/j.cageo.2022.105282.
[13]. Hong, Y. and J. Kim, Art painting detection and identification based on deep learning and image local features. Multimedia Tools and Applications, 2019. 78: p. 6513-6528
https://doi. rg/10.1007/s11042-018-6387-5
[14]. Nurhikmat, T., Implementasi deep learning untuk image classification menggunakan algoritma Convolutional Neural Network (CNN) pada citra wayang golek. 2018
https://dspace.uii.ac.id/handle/123456789/7843
[15]. Purwanto, A., Utilization of Deep Learning for Mapping Land Use Change Base on Geographic Information System: A Case Study of Liquefaction. Jurnal Penelitian Pendidikan IPA, 2023. 9(10): p. 8059-8064. https://doi.org/10.29303/jppipa.v9i10.5032
[16]. Bi, C., et al., Machine learning based fast multi-layer liquefaction disaster assessment. World Wide Web, 2019. 22: p. 1935-1950
https://doi.org/10.1007/s11280-018-0632-8
[17]. Fadhillah, M.F., et al., Integrating SAR and Optical Imagery Analysis for Liquefaction Phenomenon Identification of Post-Pohang Earthquake 2017, South Korea, Utilizing a Hybrid Deep-Learning Approach. IEEE Transactions on Geoscience and Remote Sensing, 2025.
https://doi. rg/10.1109/TGRS.2025.3550554
[18]. LONG, X., R. SUN, and T. ZHENG, Convolutional neural network-based liquefaction prediction model and interpretability analysis. Rock and Soil Mechanics, 2025. 45(9): p. 8
https://doi.org/ 10. 6599/RSM.2024.9436596
[19]. Yuan, Y., et al., A Hybrid SSA-CNN-SVM Model for Seismic-Induced Sand Liquefaction Discrimination. 2025.
https://doi.org/10.21203/rs.3.rs-7188692/v1
[20]. Ghani, S., et al., Soil categorization and liquefaction prediction using deep learning and ensemble learning algorithms. Transportation Infrastructure Geotechnology, 2025. 12(1): p. 22.
https://doi. rg/10.1007/s40515-024-00474-8
[21] .Fadhillah, M.F., et al., Liquefaction Phenomenon Identification from post-Pohang Earthquake 2017, South Korea based Integrated SAR and optical imagery data using a machine-learning approach. 2024
https://doi.org/10.21203/rs.3.rs-3849294/v1
[22]. Bai, R., et al., The Analysis of the Correlation between SPT and CPT Based on CNN-GA and Liquefaction Discrimination Research. CMES-Computer Modeling in Engineering & Sciences, 2024. 138(2).
https://doi.org/10. 2604/cmes.2023.029562
[23]. Kumar, D.R., et al., Liquefaction susceptibility using machine learning based on SPT data. Intelligent Systems with Applications, 2023. 20: p. 200281.
https://doi.org/10.1016/j.iswa.2023.200281
[24]. Şehmusoğlu, E.H., T.F. Kurnaz, and C. Erden, Estimation of soil liquefaction using artificial intelligence techniques: an extended comparison between machine and deep learning approaches. Environmental Earth Sciences, 2025. 84(5): p. 1-22 . https://doi.org/10.1007/s12665-025-12116-4
[25]. Zhang, W., et al., A deep learning approach for rapid detection of soil liquefaction using time–frequency images. Soil Dynamics and Earthquake Engineering, 2023. 166: p. 107788
https://doi.org/10.1016/j.soildyn.2023.107788
[26]. Idriss, I. and R.W. Boulanger, SPT-based liquefaction triggering procedures. Rep. UCD/CGM-10, 2010. 2: p. 4-13.
[27]. Kohestani, V., M. Hassanlourad, and A. Ardakani, Evaluation of liquefaction potential based on CPT data using random forest. Natural Hazards, 2015. 79(2): p. 1079-1089.
https://doi.org/10.1007/s11069-015-1893-5
[28]. Kayen, R., et al., Shear-wave velocity–based probabilistic and deterministic assessment of seismic soil liquefaction potential. Journal of Geotechnical and Geoenvironmental Engineering, 2013. 139(3): p. 407-419. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000743
[29]. Filali, K. and B. Sbartai, A comparative study between simplified and nonlinear dynamic methods for estimating liquefaction potential. Journal of Rock Mechanics and Geotechnical Engineering, 2017. 9(5): p. 955-966.
https://doi.org/10.1016/j.jrmge.2017.05.008
[30]. Kokusho, T., Major advances in liquefaction research by laboratory tests compared with in situ behavior. Soil Dynamics and Earthquake Engineering, 2016. 91: p. 3-22.
https://doi.org/10.1016/j.soildyn.2016.07.024
[31]. Toprak, S. and T.L. Holzer, Liquefaction potential index: field assessment. Journal of Geotechnical and Geoenvironmental Engineering, 2003. 129(4): p. 315-322.
https://doi.org/10.1061/(ASCE)1090-0241(2003)129:4(315)
[32]. Bhattacharya, S., et al., Liquefaction of soil in the Tokyo Bay area from the 2011 Tohoku (Japan) earthquake. Soil Dynamics and Earthquake Engineering, 2011. 31(11): p. 1618-1628
https://doi.org/10.1016/j.soildyn.2011.06.006
[33]. Hossain, M.B., M. Roknuzzaman, and M.M. Rahman, Liquefaction potential evaluation by deterministic and probabilistic approaches. Civil Engineering Journal, 2022. 8(7): p. 1459-1481.
http://dx.doi.org/10.28991/CEJ-2022-08-07-010
[34]. GASHAW, H., Asssessment and Evaluation of Earthquake-Induced Soil Liquefaction Potential: A Case Study in Hawassa Town. Arthitectural Engineering, 2020.
[35]. Seed, H.B., I.M. Idriss, and I. Arango, Evaluation of liquefaction potential using field performance data. Journal of geotechnical engineering, 1983. 109(3): p. 458-482.
https://doi.org/10.1061/(ASCE)0733-9410(1983)109:3(458)
[36]. Abdul Mueed, M., Machine learning algorithms-based evaluation of dynamic properties and liquefaction potential of soil. 2024, NUST
http://10.250.8.41:8080/xmlui/handle/123456789/44698
[37]. Padarian, J., B. Minasny, and A.B. McBratney, Machine learning and soil sciences: A review aided by machine learning tools. Soil, 2020. 6(1): p. 35-52. https://doi.org/10.5194/soil-6-35-2020
[38]. Hakim, W.L., et al., Land subsidence and groundwater storage assessment using ICOPS, GRACE, and susceptibility mapping in Pekalongan, Indonesia. IEEE Transactions on Geoscience and Remote Sensing, 2023. 61: p. 1-25.
https://doi.org/10.1109/TGRS.2023.3324043
[39]. Mirjalili, S., S.M. Mirjalili, and A. Hatamlou, Multi-verse optimizer: a nature-inspired algorithm for global optimization. Neural Computing and Applications, 2016. 27: p. 495-513
https://doi.org/10.1007/s00521-015-1870-7
[40]. Emami, M., Application of artifitial neural networks in pressuremeter test results. Master of Science thesis, Tarbiat Modares University, Tehran, Iran, 2009.
