结构工程师2025，Vol.41Issue(6)：62-74,13.DOI:10.15935/j.cnki.jggcs.202506.0007
基于机器学习的含再生玻璃骨料的混凝土强度预测
Machine Learning-Based Prediction of Compressive Strength of Concrete with Recycled Glass Aggregates
郁颉 1张国栋 2刘鑫 2俞可权 1余江滔1
作者信息
- 1. 同济大学结构防灾减灾工程系,上海 200092
- 2. 无锡市城市投资发展有限公司,无锡 214000
- 折叠
摘要
Abstract
To support China's dual carbon goals,this study explores the use of recycled glass aggregates as a sustainable alternative in concrete.Traditional mix design methods are often inadequate for capturing the nonlinear effects introduced by glass aggregates and involve high trial costs.Based on 331 experimental mix designs,five machine learning models(ANN,SVR,DT,RF,XGBoost)were developed to predict the 28-day compressive strength of concrete,aiming to enhance prediction efficiency and design efficiency.Results show that XGBoost and RF achieved the highest accuracy on training and testing sets(R2=99.3%and 88.5%),while ANN and RF demonstrated the best generalization on three newly developed concrete materials(R2=92.2%and 93.4%).SHAP analysis was further adopted to interpret feature contributions,confirming the models'interpretability and robustness.This study highlights the practical potential of machine learning in supporting the efficient design of eco-friendly concrete materials.关键词
再生玻璃骨料/混凝土强度/机器学习/泛化能力/SHAP模型解释Key words
recycled glass aggregates/concrete compressive strength/machine learning/generalization/SHAP analysis分类
建筑与水利引用本文复制引用
郁颉,张国栋,刘鑫,俞可权,余江滔..基于机器学习的含再生玻璃骨料的混凝土强度预测[J].结构工程师,2025,41(6):62-74,13.
