Predicting the Mechanical Properties of Concrete Containing Basalt Fiber Utilizing Optimized Hybrid Machine Learning Modeling
Buildings · 16.11.2026 · 16 (22) · 4564
DOI: 10.3390/buildings16224564 ↗Kurz gefasst
Fachartikel in der Zeitschrift Buildings, online erschienen am 16.11.2026. Verfasst von Hany A. Dahish und 1 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Basalt-fiber-reinforced concrete (BFRC) has attracted increasing attention owing to its potential to enhance concrete performance under different loading conditions. Because of the complex interactions among mixture constituents and fiber characteristics, predicting its mechanical properties can be challenging. Therefore, this study develops six individual machine learning (ML) models and a hybrid ML model based on weighted voting techniques using the six developed learners to predict both the compressive strength (CS) and tensile strength (TS) of BFRC. The individual ML models include K-nearest neighbors (KNN), random forest (RF), extra trees regressor (ETR), extreme gradient boosting (XGBoost), Light GBM (LightGBM), and gradient boosting machine (GBM). The hyperparameter optimization is performed using both particle swarm optimization (PSO) and grey wolf optimizer (GWO). Additionally, a superior hybrid ML model based on the six learners is constructed to improve the prediction accuracy of CS and TS of BFRC. Two datasets with the same mixtures containing 227 samples each were compiled from literature using eight input features: cement (C), coarse aggregate-to-fine aggregate ratio (CAG/FAG), silica fume/cement (SF/C), fly ash/cement (FA/C), water/cement ratio (W/C), superplasticizer/cement (SP/C), basalt fiber length-to-diameter ratio (F_LD), and fiber content (F_%). Analysis reveals that SF/C has the greatest influence on CS predictions, followed by W/C, CAG/FAG, F_%, and C, while TS predictions are mostly influenced by CAG/FAG, W/C, F_LD, and F_%. Finally, a graphical user interface (GUI) is developed for predicting BFRC strengths using the base learners and the hybrid model.
Zitieren
APA
Dahish, H. A., & Alenezi, A. M. (2026). Predicting the Mechanical Properties of Concrete Containing Basalt Fiber Utilizing Optimized Hybrid Machine Learning Modeling. Buildings, 16(22), 4564. https://doi.org/10.3390/buildings16224564
Vancouver
Dahish HA, Alenezi AM. Predicting the Mechanical Properties of Concrete Containing Basalt Fiber Utilizing Optimized Hybrid Machine Learning Modeling. Buildings. 2026;16(22):4564. doi:10.3390/buildings16224564
BibTeX
@article{Dahish2026,
author = {Dahish, Hany A. and Alenezi, Ahmad Mfarreh},
title = {Predicting the Mechanical Properties of Concrete Containing Basalt Fiber Utilizing Optimized Hybrid Machine Learning Modeling},
journal = {Buildings},
year = {2026},
volume = {16},
number = {22},
pages = {4564},
doi = {10.3390/buildings16224564}
}
EUVIMED