Forecasting early-age thermal cracking risk in structural concrete with noisy data
Transport and Communications Science Journal · 06.05.2025 · 76 (5) · 727-738
DOI: 10.47869/tcsj.76.5.7 ↗Kurz gefasst
Fachartikel in der Zeitschrift Transport and Communications Science Journal, online erschienen am 06.05.2025. Verfasst von Tiến Nguyễn Duy und 3 weiteren Autor:innen.
Abstract (Originalsprache)
Numerical methods have demonstrated significant advantages in simulating and predicting early-age thermal cracking in concrete (commonly in mass concrete), as traditional measurements are often time-consuming and costly. This study integrates numerical methods (to determine input parameters) with machine learning techniques to predict the early-age cracking potential of structural concrete blocks. A dataset of 759 samples on the early-age thermal cracking of bridge piers was collected by varying the height and width of the piers. Subsequently, this data with 10% noise was used to train the LR, SVR, RF, KNN, Adaboost, XGBoost, and CatBoost models. After training, the resulting models were used to predict the outcomes of the 759 samples with 0% noise. With a prediction accuracy of R² = 0,9618 for the cracking index, the CatBoost model performs best, even maintaining high accuracy when the training data contains up to 10% noise. This study contributes to demonstrating the applicability of machine learning methods in reducing computational time and enhancing prediction accuracy of early-age thermal cracking in concrete, even in cases where input data quality is low (noisy)
Zitieren
APA
Duy, T. N., Anh, T. Đ., Việt, H. H., & Hùng, C. L. (2025). Forecasting early-age thermal cracking risk in structural concrete with noisy data. Transport and Communications Science Journal, 76(5), 727-738. https://doi.org/10.47869/tcsj.76.5.7
Vancouver
Duy TN, Anh TĐ, Việt HH, Hùng CL. Forecasting early-age thermal cracking risk in structural concrete with noisy data. Transport and Communications Science Journal. 2025;76(5):727-738. doi:10.47869/tcsj.76.5.7
BibTeX
@article{Duy2025,
author = {Duy, Tiến Nguyễn and Anh, Tú Đỗ and Việt, Hải Hoàng and Hùng, Cường Lưu},
title = {Forecasting early-age thermal cracking risk in structural concrete with noisy data},
journal = {Transport and Communications Science Journal},
year = {2025},
volume = {76},
number = {5},
pages = {727--738},
doi = {10.47869/tcsj.76.5.7}
}
EUVIMED