تولید داده‌های مصنوعی برای پیش‌بینی خواص بتن: نوآوری‌ها و کاربردها در مهندسی محاسباتی
محورهای موضوعی : مجله فناوری اطلاعات در طراحی مهندسیسید ایمان غفوریان حیدری 1 , مجید صافحیان 2 * , فرامرز مودی 3 , شبنم شادرو 4
1 - دانشگاه آزاد اسلامی واحد مشهد
2 - استاد یار دانشگاه علوم و تحقیقات
3 - استادیار مرکز تحقیقات فناوری و دوام بتن، دانشگاه فناوری امیرکبیر، تهران، ایران
4 - گروه مهندسی کامپیوتر، واحد مشهد، دانشگاه آزاد اسلامی، مشهد، ایران
کلید واژه: کلمات کلیدی: تولید داده‌های مصنوعی, یادگیری ماشین, مقاومت فشاری بتن, K-Means, TGAN,
چکیده مقاله :
چالش‌های موجود در دسترسی به داده‌های آزمایشگاهی دقیق، استفاده از داده‌های مصنوعی را به‌عنوان راه‌حلی کارآمد برای بهبود پیش‌بینی مقاومت فشاری بلندمدت بتن تحت تأثیر مواد افزودنی شیمیایی مطرح کرده است. در این پژوهش، روش‌های پیشرفته برای تولید داده‌های مصنوعی و تحلیل آن‌ها به‌منظور پیش‌بینی مقاومت فشاری بتن بررسی شده است. داده‌های مصنوعی در سطوح مختلف تولید و ارزیابی شدند تا تعداد بهینه برای دستیابی به بهترین نتایج شناسایی شود. نتایج نشان داد که تولید داده‌های مصنوعی می‌تواند دقت مدل‌های پیش‌بینی را به‌طور چشمگیری افزایش دهد و همچنین نیاز به انجام آزمایش‌های فیزیکی را کاهش دهد.در این مطالعه از ترکیب داده‌های آزمایشگاهی و داده‌های مصنوعی استفاده شد و تأثیر مثبت آن بر بهبود عملکرد مدل‌های پیش‌بینی مورد تأیید قرار گرفت. روش پیشنهادی در این پژوهش امکان تولید داده‌های مصنوعی مرتبط با شرایط واقعی را فراهم کرده و باعث صرفه‌جویی در منابع، کاهش هزینه‌ها و کاهش تأثیرات زیست‌محیطی شده است. یافته‌ها همچنین نشان دادند که مقاومت فشاری بتن طی دوره سه‌ساله به‌طور قابل توجهی کاهش می‌یابد که می‌تواند عمر مفید سازه‌ها را کاهش داده و هزینه‌های تعمیرات را افزایش دهد.
The challenges associated with obtaining accurate experimental data have underscored the utility of synthetic data as an effective solution for enhancing the long-term prediction of compressive strength in concrete influenced by chemical admixtures. This study explores advanced methodologies for the generation and analysis of synthetic data aimed at predicting the compressive strength of concrete. Synthetic data were generated and assessed at various levels to identify the optimal quantity required to achieve the most favorable outcomes. The results indicated that the generation of synthetic data can significantly improve the accuracy of predictive models while concurrently diminishing the reliance on physical experiments.
In this research, a hybrid approach utilizing both experimental and synthetic data was employed, with its beneficial effects on the performance of predictive models being substantiated. The proposed methodology facilitates the generation of synthetic data that accurately simulates real-world conditions, resulting in resource conservation, cost savings, and a reduction in environmental impact. Additionally, the findings revealed a significant decline in the compressive strength of concrete over a three-year period, which has implications for the service life of structures and may lead to increased maintenance costs.
[1] A. Harapin, M. Jurišić, N. Bebek, and M. Sunara, “Long-Term Effects in Structures: Background and Recent Developments,” Applied Sciences (Switzerland), vol. 14, no. 6. Multidisciplinary Digital Publishing Institute (MDPI), 2024. doi: 10.3390/app14062352.
[2] M. R. Garcez, A. B. Rohden, and L. G. G. de Godoy, “The role of concrete compressive strength on the service life and life cycle of a RC structure: Case study,” J. Clean. Prod., vol. 172, pp. 27–38, 2018, doi: 10.1016/j.jclepro.2017.10.153.
[3] K. Bedada, A. Nyabuto, I. Kınotı, and J. Marangu, “Review on advances in bio-based admixtures for concrete,” J. Sustain. Constr. Mater. Technol., vol. 8, no. 4, pp. 344–367, 2023.
[4] S. Zeng, X. Wang, L. Hua, M. Altayeb, Z. Wu, and F. Niu, “Prediction of compressive strength of FRP-confined concrete using machine learning: A novel synthetic data driven framework,” J. Build. Eng., vol. 94, p. 109918, 2024, doi: https://doi.org/10.1016/j.jobe.2024.109918.
[5] R. Polo-Mendoza, G. Martinez-Arguelles, R. Peñabaena-Niebles, and J. Duque, “Development of a Machine Learning (ML)-Based Computational Model to Estimate the Engineering Properties of Portland Cement Concrete (PCC),” Arab. J. Sci. Eng., no. Ml, 2024, doi: 10.1007/s13369-024-08794-0.
[6] H. A. T. Nguyen, D. H. Pham, and Y. Ahn, “Effect of Data Augmentation Using Deep Learning on Predictive Models for Geopolymer Compressive Strength,” Appl. Sci., vol. 14, no. 9, 2024, doi: 10.3390/app14093601.
[7] N. Chen et al., “Virtual mix design: Prediction of compressive strength of concrete with industrial wastes using deep data augmentation,” Constr. Build. Mater., vol. 323, no. January, 2022, doi: 10.1016/j.conbuildmat.2022.126580.
[8] T. Aziz, H. Aziz, S. Mahapakulchai, and C. Charoenlarpnopparut, “Optimizing compressive strength prediction using adversarial learning and hybrid regularization,” Sci. Rep., vol. 14, no. 1, pp. 1–14, 2024, doi: 10.1038/s41598-024-69434-z.
[9] Q. Tian et al., “Compressive strength of waste-derived cementitious composites using machine learning,” Rev. Adv. Mater. Sci., vol. 63, no. 1, 2024, doi: 10.1515/rams-2024-0008.
[10] Y. El Khessaimi et al., “The Effectiveness of Data Augmentation in Compressive Strength Prediction of Calcined Clay Cements Using Linear Regression Learning,” NanoWorld J., vol. 9, 2023, doi: 10.17756/nwj.2023-s2-054.
[11] A. Marani, A. Jamali, and M. L. Nehdi, “Predicting ultra-high-performance concrete compressive strength using tabular generative adversarial networks,” Materials (Basel)., vol. 13, no. 21, pp. 1–24, 2020, doi: 10.3390/ma13214757.
[12] J. Pachouly, S. Ahirrao, K. Kotecha, G. Selvachandran, and A. Abraham, “A systematic literature review on software defect prediction using artificial intelligence: Datasets, Data Validation Methods, Approaches, and Tools,” Eng. Appl. Artif. Intell., vol. 111, no. November 2021, p. 104773, 2022, doi: 10.1016/j.engappai.2022.104773.
[13] M. Shariati, D. J. Armaghani, M. Khandelwal, J. Zhou, and M. Khorami, “Assessment of Longstanding Effects of Fly Ash and Silica Fume on the Compressive Strength of Concrete Using Extreme Learning Machine and Artificial Neural Network,” J. Adv. Eng. Comput., vol. 5, no. 1, p. 50, Mar. 2021, doi: 10.25073/jaec.202151.308.
[14] L. Jin, W. Yu, D. Li, and X. Du, “Numerical and theoretical investigation on the size effect of concrete compressive strength considering the maximum aggregate size,” Int. J. Mech. Sci., vol. 192, p. 106130, 2021.
[15] A. International, “Standard Specification for Concrete Aggregates- ASTM C33 / C33M-23,” 2023. [Online]. Available: 10.1520/C0033_C0033M-23
[16] N. Zeminian, G. Guarino, and Q. Xu, “Chemical admixtures for the optimization of the AAC production,” Ce/papers, vol. 2, no. 4, pp. 235–240, 2018.
[17] A. International, “Standard Practice for Making and Curing Concrete Test Specimens in the Laboratory: ASTM C192/C192M-24,” West Conshohocken, PA, 2024, 2024. doi: 10.1520/C0192_C0192M-24.
[18] V. Çetin and O. Yıldız, “A comprehensive review on data preprocessing techniques in data analysis,” Pamukkale Univ. J. Eng. Sci., vol. 28, no. 2, pp. 299–312, 2022, doi: 10.5505/pajes.2021.62687.
[19] I. Nunez, A. Marani, M. Flah, and M. L. Nehdi, “Estimating compressive strength of modern concrete mixtures using computational intelligence: A systematic review,” Constr. Build. Mater., vol. 310, p. 125279, 2021, doi: https://doi.org/10.1016/j.conbuildmat.2021.125279.
[20] L. Xu and K. Veeramachaneni, “Synthesizing Tabular Data using Generative Adversarial Networks,” 2018, [Online]. Available: http://arxiv.org/abs/1811.11264
[21] K. P. Sinaga and M. S. Yang, “Unsupervised K-means clustering algorithm,” IEEE Access, vol. 8, pp. 80716–80727, 2020, doi: 10.1109/ACCESS.2020.2988796.
[22] K. Miok, D. Nguyen-Doan, D. Zaharie, and M. Robnik-Šikonja, “Generating data using Monte Carlo dropout,” in 2019 IEEE 15th International Conference on Intelligent Computer Communication and Processing (ICCP), IEEE, 2019, pp. 509–515.
[23] D. Kaur et al., “Application of Bayesian networks to generate synthetic health data,” J. Am. Med. Informatics Assoc., vol. 28, no. 4, pp. 801–811, 2021.
[24] S. J. Alghamdi, “Prediction of Concrete’s Compressive Strength via Artificial Neural Network Trained on Synthetic Data,” Eng. Technol. Appl. Sci. Res., vol. 13, no. 6, pp. 12404–12408, 2023.
[25] Y. Lu, M. Shen, H. Wang, X. Wang, C. van Rechem, and W. Wei, “Machine learning for synthetic data generation: a review,” arXiv Prepr. arXiv2302.04062, 2023.
[26] M. Z. Naser and A. H. Alavi, “Error metrics and performance fitness indicators for artificial intelligence and machine learning in engineering and sciences,” Archit. Struct. Constr., vol. 3, no. 4, pp. 499–517, 2023.
[27] A. N. Beskopylny et al., “Prediction of the Compressive Strength of Vibrocentrifuged Concrete Using Machine Learning Methods,” Buildings, vol. 14, no. 2, pp. 1–20, 2024, doi: 10.3390/buildings14020377.
[28] F. Barreto, L. Moharkar, M. Shirodkar, V. Sarode, S. Gonsalves, and A. Johns, “Generative Artificial Intelligence: Opportunities and Challenges of Large Language Models BT - Intelligent Computing and Networking,” V. E. Balas, V. B. Semwal, and A. Khandare, Eds., Singapore: Springer Nature Singapore, 2023, pp. 545–553.
[29] L. Xu, M. Skoularidou, A. Cuesta-Infante, and K. Veeramachaneni, “Modeling tabular data using conditional gan,” Adv. Neural Inf. Process. Syst., vol. 32, 2019.
[30] L. Zhang, A. Gonzalez-Garcia, J. Van De Weijer, M. Danelljan, and F. S. Khan, “Synthetic data generation for end-to-end thermal infrared tracking,” IEEE Trans. Image Process., vol. 28, no. 4, pp. 1837–1850, 2018.
[31] M. Saito, E. Matsumoto, and S. Saito, “Temporal generative adversarial nets with singular value clipping,” in Proceedings of the IEEE international conference on computer vision, 2017, pp. 2830–2839.
[32] P. Ziolkowski and M. Niedostatkiewicz, “Machine learning techniques in concrete mix design,” Materials (Basel)., vol. 12, no. 8, 2019, doi: 10.3390/ma12081256.
[33] A. L. Kaminsky, Y. Wang, and K. Pant, “An efficient batch K-fold cross-validation voronoi adaptive sampling technique for global surrogate modeling,” J. Mech. Des., vol. 143, no. 1, p. 11706, 2021.
[34] K. Phinzi, D. Abriha, and S. Szabó, “Classification efficacy using k-fold cross-validation and bootstrapping resampling techniques on the example of mapping complex gully systems,” Remote Sens., vol. 13, no. 15, p. 2980, 2021.
[35] J. S. Chou and A. D. Pham, “Enhanced artificial intelligence for ensemble approach to predicting high performance concrete compressive strength,” Constr. Build. Mater., vol. 49, pp. 554–563, 2013, doi: 10.1016/j.conbuildmat.2013.08.078.
[36] H. Song et al., “Predicting the compressive strength of concrete with fly ash admixture using machine learning algorithms,” Constr. Build. Mater., vol. 308, p. 125021, 2021.
[37] “Anaconda, Anaconda,” World’s Most Pop. Data Sci. Platf., 2019.
[38] T. W. M. P. D. S. Platform, “Matlab R2021a,” World’s Most Pop. Data Sci. Platf., [Online]. Available: The World’s Most Popular Data Science Platform
[39] X. Hu, B. Li, Y. Mo, and O. Alselwi, “Progress in artificial intelligence-based prediction of concrete performance,” J. Adv. Concr. Technol., vol. 19, no. 8, pp. 924–936, 2021.
[40] H. Naderpour, A. H. Rafiean, and P. Fakharian, “Compressive strength prediction of environmentally friendly concrete using artificial neural networks,” J. Build. Eng., vol. 16, no. January, pp. 213–219, 2018, doi: 10.1016/j.jobe.2018.01.007.
[41] X. Dong, Y. Liu, and J. Dai, “Application of Fully Connected Neural Network‐Based PyTorch in Concrete Compressive Strength Prediction,” Adv. Civ. Eng., vol. 2024, no. 1, p. 8048645, 2024.
[42] D. H. de Bem, D. P. B. Lima, and R. A. Medeiros-Junior, “Effect of chemical admixtures on concrete’s electrical resistivity,” Int. J. Build. Pathol. Adapt., vol. 36, no. 2, pp. 174–187, 2018, doi: 10.1108/IJBPA-11-2017-0058.
[43] M. Saridemir, I. B. Topçu, F. Özcan, and M. H. Severcan, “Prediction of long-term effects of GGBFS on compressive strength of concrete by artificial neural networks and fuzzy logic,” Constr. Build. Mater., vol. 23, no. 3, pp. 1279–1286, Mar. 2009, doi: 10.1016/j.conbuildmat.2008.07.021.