Prediction of concrete and FRC properties at high temperature using machine and deep learning: A review of recent advances and future perspectives

Journal of Building Engineering, Год журнала: 2023, Номер 83, С. 108369 - 108369

Опубликована: Дек. 29, 2023

Язык: Английский

---

Compressive strength of concrete material using machine learning techniques

Cleaner Engineering and Technology, Год журнала: 2023, Номер 15, С. 100661 - 100661

Опубликована: Июль 20, 2023

Significant efforts have been made to improve the strength of concrete by utilizing industrial waste like Fly Ash as a partial replacement cement in concrete. However, predicting compressive is one challenging tasks since it affected several factors such shape and size aggregates, water-cement ratio. The paper presents study on various investigation machine learning (ML) algorithms estimate (CS) containing fly ash (FA). research also aims compare accuracy different ML models, including non-ensemble models (Multiple Linear Regressor, Support Vector Regressor) ensemble (AdaBoost Random Forest Regression, XGBoost Bagging Regressor), CS with focus identifying most accurate estimation method. For this purpose, dataset 633 experimental results wide range values, ranging from 6.27 MPa 79.99 MPa, was collected existing literature validated using statistical analysis. primary input parameters for included quantities cement, fine aggregate (FA), coarse aggregates (CA), water content, percentage superplasticizer, curing days, output. Performance evaluation conducted performance indices, MAE, MSE, R2, MAPE, RMSE, a20-index, assess reliability. comparison reveals that Regressor reliable model, demonstrating highest coefficient determination (R2) 0.95, a-20 index 0.913, lowest RMSE value 3.06 MAE 2.13 while Multiple LR model least method R2 equal 0.52, 0.433, 9.40 7.68 MPa. Additionally, provide deeper insights into relationship between CS, sensitivity parametric analysis were employed, enabling comprehensive understanding impact other prediction. From study, observed age essential feature, followed water, information gain values 32.91, 23.50, 15.10, respectively. highlights effectiveness techniques, particularly accurately estimating Furthermore, offers researchers faster more cost-effective means evaluating effect estimation, avoiding need time-consuming costly studies.

Язык: Английский

---

An explainable machine learning approach to predict the compressive strength of graphene oxide-based concrete

Construction and Building Materials, Год журнала: 2024, Номер 449, С. 138346 - 138346

Опубликована: Сен. 17, 2024

Язык: Английский

---

Predicting carbonation depth of concrete using a hybrid ensemble model

Journal of Building Engineering, Год журнала: 2023, Номер 76, С. 107320 - 107320

Опубликована: Июль 12, 2023

Язык: Английский

---

A comparative study of machine learning algorithms for the prediction of compressive strength of rice husk ash-based concrete

Natural Hazards, Год журнала: 2023, Номер 118(1), С. 209 - 238

Опубликована: Май 9, 2023

Язык: Английский

---

The use of treated desert sand in sustainable concrete: A mechanical and microstructure study

Journal of Building Engineering, Год журнала: 2023, Номер 79, С. 107843 - 107843

Опубликована: Сен. 26, 2023

Язык: Английский

---

Comparative analysis of various machine learning algorithms to predict strength properties of sustainable green concrete containing waste foundry sand

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Июнь 25, 2024

Abstract The use of waste foundry sand (WFS) in concrete production has gained attention as an eco-friendly approach to reduction and enhancing cementitious materials. However, testing the impact WFS through experiments is costly time-consuming. Therefore, this study employs machine learning (ML) models, including support vector regression (SVR), decision tree (DT), AdaBoost regressor (AR) ensemble model predict properties accurately. Moreover, SVR was employed conjunction with three robust optimization algorithms: firefly algorithm (FFA), particle swarm (PSO), grey wolf (GWO), construct hybrid models. Using 397 experimental data points for compressive strength (CS), 146 elastic modulus (E), 242 split tensile (STS), models were evaluated statistical metrics interpreted using SHapley Additive exPlanation (SHAP) technique. SVR-GWO demonstrated exceptional accuracy predicting (WFSC) characteristics. exhibited correlation coefficient values (R) 0.999 CS E, 0.998 STS. Age found be a significant factor influencing WFSC properties. also comparable prediction model. In addition, SHAP analysis revealed optimal content input variables mix. Overall, showed compared individual application these sophisticated soft computing techniques holds potential stimulate widespread adoption sustainable production, thereby fostering bolstering environmentally conscious construction practices.

Язык: Английский

---

Interpretable predictive modeling, sustainability assessment, and cost analysis of cement-based composite containing secondary raw materials

Construction and Building Materials, Год журнала: 2025, Номер 473, С. 140924 - 140924

Опубликована: Март 28, 2025

Язык: Английский

---

Predicting the elastic modulus of normal and high strength concretes using hybrid ANN-PSO

Materials Today Proceedings, Год журнала: 2023, Номер unknown

Опубликована: Март 1, 2023

In the design and analysis stages, modulus of elasticity plays a crucial role in influencing lateral deflection reinforced concrete structure. The elastic (EM) is affected by wide variety parameters, some which are compressive capacity sample, its age, type aggregate used, cement loading rate, size test sample. A precise estimation accordance with accepted guidelines challenging process requiring specialized protocols continuous strain monitoring. Intelligent systems, particularly artificial neural networks, now widely used to be considered general efficient tools for applied research. As brain does, networks information similarly. To solve particular problem, large number closely connected processing components work simultaneously. identify weighting factors, other approaches may utilized; however, particle swarm optimization (PSO) algorithm was chosen this belongs field intelligence method global minimization that can problems where answer point or surface n-dimensional space. An innovative hybrid network-based model determining EM samples presented study. similar manner conventional regulations, proposed models have been determined using strength parameter simply. confirm correctness model, thorough comparison conducted between laboratory values, findings from research, outcomes previously established relationships. For comparison, code relations such as ACI 318, 363, FIB, NS 3473 were used. suggested operates excellently helps determine concrete, shown results.

Язык: Английский

---

Soft computing-based prediction models for compressive strength of concrete

Case Studies in Construction Materials, Год журнала: 2023, Номер 19, С. e02321 - e02321

Опубликована: Июль 22, 2023

The complexity of concrete's composition makes it difficult to predict its compressive strength, which is a highly valuable and desired characteristic. Traditional methods for prediction are expensive time-consuming, resulting in limited data availability. However, modern soft-computing models have emerged as reliable solution accurately forecasting strength. research proposes novel Deep Neural Network (DNN), Multivariate Adaptive Regression Splines (MARS) Extreme Learning Machine (ELM) based machine learning (ML) the strength concrete added with various proportions fly ash silica fume. For this purpose, dataset 144 trials, having 8 input parameters taken from literature. performance confirmed using statistical parameters. Rank Analysis reveals that DNN best-performing model (Rank =52, RTR2=0.983 RTs2=0.954), closely followed by MARS =38, RTR2=0.974 RTs2=0.956); while ELM lags behind other two counterparts. results further an error matrix, external validation AIC criteria. visual interpretation provided Taylor diagram. has edge over terms providing user-friendly solution.

Язык: Английский

---