Application of machine learning for predicting adverse thermal gradients in reinforced concrete box-girder bridges
Transport and Communications Science Journal · 02.06.2025 · 76 (8) · 1123-1137
DOI: 10.47869/tcsj.76.8.7 ↗Kurz gefasst
Fachartikel in der Zeitschrift Transport and Communications Science Journal, online erschienen am 02.06.2025. Verfasst von Cảnh Lê Minh und 2 weiteren Autor:innen.
Abstract (Originalsprache)
During the operational phase, reinforced concrete (RC) box-girder bridges are significantly influenced by environmental temperature variations. Temperature differentials among different parts of the cross-section generate considerable thermal stresses, reducing structural durability and lifespan. Recent studies highlight substantial discrepancies between vertical thermal gradients measured in the field and the standard values prescribed by the Vietnamese standard TCVN 11823:2017. This study analyzes nearly three years of continuous temperature measurement data from three representative bridges located in distinct climatic regions of Vietnam: Vinh Thinh Bridge (Northern region), Quan Hau Bridge (Central region), and Go Gang Bridge (Southern region). To enhance the accuracy of thermal effect assessments, supervised machine learning models including Linear Regression, Boosted Tree, Random Forest, XGBoost, and Artificial Neural Network (ANN) were developed to predict critical thermal gradient values. Among these models, the ANN demonstrated superior performance, achieving coefficient of determination (R2) above 0,70 for both MAX-Gy and MIN-Gy. This result indicates strong predictive capabilities and effectiveness in capturing nonlinear relationships between meteorological conditions and the geometric characteristics of bridge cross-sections. The outcomes provide a robust scientific foundation for the design, evaluation, and maintenance of RC box-girder bridges in tropical climates, ultimately contributing to increased durability and operational efficiency
Zitieren
APA
Minh, C. L., Đức, C. N., & Thành, S. L. (2025). Application of machine learning for predicting adverse thermal gradients in reinforced concrete box-girder bridges. Transport and Communications Science Journal, 76(8), 1123-1137. https://doi.org/10.47869/tcsj.76.8.7
Vancouver
Minh CL, Đức CN, Thành SL. Application of machine learning for predicting adverse thermal gradients in reinforced concrete box-girder bridges. Transport and Communications Science Journal. 2025;76(8):1123-1137. doi:10.47869/tcsj.76.8.7
BibTeX
@article{Minh2025,
author = {Minh, Cảnh Lê and Đức, Chuyên Nguyễn and Thành, Sang Lê},
title = {Application of machine learning for predicting adverse thermal gradients in reinforced concrete box-girder bridges},
journal = {Transport and Communications Science Journal},
year = {2025},
volume = {76},
number = {8},
pages = {1123--1137},
doi = {10.47869/tcsj.76.8.7}
}
EUVIMED