Concrete is one of the most widely used materials in hydraulic engineering, and its compressive strength is a crucial indicator for assessing the safety and durability of water-related structures. Traditional machine learning models have demonstrated potential in predicting concrete performance; however, single model approaches often suffer from overfitting and limited accuracy.
This study proposes a novel hybrid predictive framework integrating multiple soft computing techniques and metaheuristic optimization algorithms to improve the prediction of hydraulic concrete compressive strength. Classic machine learning models—support vector machine (SVM), random forest (RF), Gaussian process regression (GPR), and artificial neural network (ANN)—were selected as base learners, and their optimal hyperparameters were tuned using an improved gray wolf optimization (GWO) algorithm. In the second stage, a lightweight gradient boosting machine (LightGBM) was employed as a meta-learner in a stacking ensemble structure to integrate the outputs of base models and enhance generalization performance.
A dataset of 1050 samples was established using both open-source and experimental data of hydraulic concrete specimens. The ensemble model achieved a regression coefficient (R²) of 0.9329, a mean absolute error (MAE) of 2.7695, and a mean square error (MSE) of 4.0891, outperforming all single models and other ensemble methods. Feature importance analysis identified cement dosage, coarse-to-fine aggregate ratio, water–cement ratio and curing age as the dominant factors influencing strength.
This integrated hybrid model demonstrates excellent robustness and generalization capacity for structural damage assessment and life prediction of hydraulic concrete. The proposed framework offers a powerful decision-support tool for sustainable design and intelligent monitoring of water-related infrastructures.
Yang Yang is a postgraduate student in the School of Civil and Environmental Engineering at the University of New South Wales (UNSW), Australia. He earned his Bachelor of Engineering degree in Civil Engineering from Ningbo University in 2022. His research interests include concrete materials, structural mechanics, and engineering optimization techniques. He previously worked on projects in tunnel construction and bridge reinforcement during internships with China Railway Tunnel Group and CCCC Third Harbor Engineering Co., Ltd.
Copyright 2024 Mathews International LLC All Rights Reserved