Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization
Basalt fiber reinforced concrete (BFRC) is widely used in bridges, tunnels, and seismic-resistant structures due to its excellent crack resistance, durability, and environmental benefits. However, multiple factors with complex nonlinear behavior affect its compressive and splitting tensile strength....
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2025-01-01
|
| Series: | Materials Research Express |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2053-1591/ade11d |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849333986744074240 |
|---|---|
| author | Dalian Bai Junwen Chen Jialiang Wang Cunpeng Liu |
| author_facet | Dalian Bai Junwen Chen Jialiang Wang Cunpeng Liu |
| author_sort | Dalian Bai |
| collection | DOAJ |
| description | Basalt fiber reinforced concrete (BFRC) is widely used in bridges, tunnels, and seismic-resistant structures due to its excellent crack resistance, durability, and environmental benefits. However, multiple factors with complex nonlinear behavior affect its compressive and splitting tensile strength. Traditional experiments are time-consuming and limited in capturing this behavior. This study applies the Random Forest (RF) algorithm combined with hyperparameter optimization methods including Grid Search (GS), Random Search (RS), Bayesian Optimization (BO), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Optuna Optimization (OO) to build a predictive model. A systematic comparison of different optimization methods was conducted using ten-fold cross-validation, and the results were based on 314 compressive strength test samples and 293 splitting tensile strength test samples. The results indicate that the BO achieved the highest accuracy and efficiency, while the OO excelled in computation time. Feature importance and SHAP (SHapley Additive exPlanations) analysis identified the cement content (400 to 650 kg m ^−3 ) as the key factor. A moderate fiber length (15 to 20 mm) and fiber dosage (0.1% to 0.15%) contribute to improving the compressive strength, while an appropriate fiber diameter (0.015 mm), fiber length (15 to 20 mm), and fiber dosage (0.15% to 0.25%) effectively enhance the splitting tensile strength. |
| format | Article |
| id | doaj-art-1fa994e36cee4f94b9569f8b493553f5 |
| institution | Kabale University |
| issn | 2053-1591 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Materials Research Express |
| spelling | doaj-art-1fa994e36cee4f94b9569f8b493553f52025-08-20T03:45:41ZengIOP PublishingMaterials Research Express2053-15912025-01-0112606570210.1088/2053-1591/ade11dMechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimizationDalian Bai0https://orcid.org/0009-0000-2562-4272Junwen Chen1Jialiang Wang2Cunpeng Liu3School of Civil and Surveying Engineering at Nanning Branch, Guilin University of Technology , Nanning, 530001, People’s Republic of ChinaSchool of Civil and Surveying Engineering at Nanning Branch, Guilin University of Technology , Nanning, 530001, People’s Republic of ChinaSchool of Civil and Surveying Engineering at Nanning Branch, Guilin University of Technology , Nanning, 530001, People’s Republic of ChinaSchool of Civil and Surveying Engineering at Nanning Branch, Guilin University of Technology , Nanning, 530001, People’s Republic of ChinaBasalt fiber reinforced concrete (BFRC) is widely used in bridges, tunnels, and seismic-resistant structures due to its excellent crack resistance, durability, and environmental benefits. However, multiple factors with complex nonlinear behavior affect its compressive and splitting tensile strength. Traditional experiments are time-consuming and limited in capturing this behavior. This study applies the Random Forest (RF) algorithm combined with hyperparameter optimization methods including Grid Search (GS), Random Search (RS), Bayesian Optimization (BO), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Optuna Optimization (OO) to build a predictive model. A systematic comparison of different optimization methods was conducted using ten-fold cross-validation, and the results were based on 314 compressive strength test samples and 293 splitting tensile strength test samples. The results indicate that the BO achieved the highest accuracy and efficiency, while the OO excelled in computation time. Feature importance and SHAP (SHapley Additive exPlanations) analysis identified the cement content (400 to 650 kg m ^−3 ) as the key factor. A moderate fiber length (15 to 20 mm) and fiber dosage (0.1% to 0.15%) contribute to improving the compressive strength, while an appropriate fiber diameter (0.015 mm), fiber length (15 to 20 mm), and fiber dosage (0.15% to 0.25%) effectively enhance the splitting tensile strength.https://doi.org/10.1088/2053-1591/ade11dbasalt fiber reinforced concretecompressive strengthtensile strengthrandom foresthyperparameter optimization |
| spellingShingle | Dalian Bai Junwen Chen Jialiang Wang Cunpeng Liu Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization Materials Research Express basalt fiber reinforced concrete compressive strength tensile strength random forest hyperparameter optimization |
| title | Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| title_full | Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| title_fullStr | Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| title_full_unstemmed | Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| title_short | Mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| title_sort | mechanical performance prediction of basalt fiber reinforced concrete based on random forest and hyperparameter optimization |
| topic | basalt fiber reinforced concrete compressive strength tensile strength random forest hyperparameter optimization |
| url | https://doi.org/10.1088/2053-1591/ade11d |
| work_keys_str_mv | AT dalianbai mechanicalperformancepredictionofbasaltfiberreinforcedconcretebasedonrandomforestandhyperparameteroptimization AT junwenchen mechanicalperformancepredictionofbasaltfiberreinforcedconcretebasedonrandomforestandhyperparameteroptimization AT jialiangwang mechanicalperformancepredictionofbasaltfiberreinforcedconcretebasedonrandomforestandhyperparameteroptimization AT cunpengliu mechanicalperformancepredictionofbasaltfiberreinforcedconcretebasedonrandomforestandhyperparameteroptimization |