Interpretable machine‐learning models for predicting creep recovery of concrete DOI
Shengqi Mei, Xiaodong Liu, Xingju Wang

и другие.

Structural Concrete, Год журнала: 2024, Номер unknown

Опубликована: Окт. 16, 2024

Abstract Creep recovery of concrete is essential for accurately assessing the performance structures over service time. Existing creep models exhibit low accuracy, and influencing factors remain inadequately elucidated. In this paper, interpretable machine learning (ML) techniques were employed to develop a prediction model recovery. Several ML selected including random forest (RF), support vector regression (SVR), extreme gradient boosting (XGBoost) light (LGBM). order maximize sample size dataset, 109 sets data collected from existing literatures training. Feature selection utilized determine input parameters models, 12 variables selected. The fine‐tuned using Bayesian optimization techniques. To ensure reliability 10‐fold cross‐validation splitting implemented. results indicate that exhibited higher accuracy compared model. Among these LGBM demonstrated superior efficiency stability (with R 2 = 0.993, 0.978, 0.973 training, testing, validation sets, respectively). Shapley additive explanations (SHAP) interpret significance each parameter on prediction. Duration after unloading, stress magnitude, ambient relative humidity main feature Upon comparing factors, it was discerned there exists distinct difference between concrete.

Язык: Английский

Optimizing high-strength concrete compressive strength with explainable machine learning DOI Creative Commons
Sanjog Chhetri Sapkota,

Christina Panagiotakopoulou,

Dipak Dahal

и другие.

Multiscale and Multidisciplinary Modeling Experiments and Design, Год журнала: 2025, Номер 8(3)

Опубликована: Фев. 3, 2025

Язык: Английский

Процитировано

3
Predicting Compressive Strength of Oil Well Cement Slurries: Novel Moduli‐Based Analysis of Chemical Composition at Different Temperature Condition DOI Open Access
Mohammed A. Jamal, Ahmed Salih Mohammed, Jagar A. Ali

и другие.

The Structural Design of Tall and Special Buildings, Год журнала: 2025, Номер 34(2)

Опубликована: Янв. 30, 2025

ABSTRACT This study evaluates the impact of cement chemical composition on compressive strength (CS) slurries, utilizing silica fume (SF) and fly ash (FA) as additional materials. A comprehensive analysis was conducted 317 datasets from literature, focusing factors including silicon dioxide (SiO₂), aluminum oxide (Al₂O₃), calcium (CaO), iron (Fe₂O₃), water‐to‐binder (w/b) ratio, SF FA content, well curing time temperature. The research presents three geochemical moduli, namely, silicate modulus (SM), aluminate (AM), hydraulic (HM), to assess forecast CS. investigation full quadratic (FQ) cubic (CUB) models underscores precision prediction corroborated by statistical metrics, such scatter index (SI), root mean squared error (RMSE), correlation coefficient ( R 2 ). Univariate, bivariate, multivariate evaluations indicate that SM, AM, HM significantly decrease input parameters while preserving or enhancing model accuracy. ideal replacement percentages for maximize were determined be 14.6% 11.6%, respectively. optimal values 2.62, 1.38, 2.21, results establish a solid framework optimizing formulations, presenting sustainable alternatives improved mechanical performance decreased material consumption in oil cementing building applications.

Язык: Английский

Процитировано

1
EVALUATING THE MECHANICAL AND DURABILITY PROPERTIES OF SUSTAINABLE LIGHTWEIGHT CONCRETE INCORPORATING THE VARIOUS PROPORTIONS OF WASTE PUMICE AGGREGATE DOI Creative Commons
Hafiz Muhammad Shahzad Aslam, Atteq ur Rehman, Kennedy C. Onyelowe

и другие.

Results in Engineering, Год журнала: 2024, Номер unknown, С. 103496 - 103496

Опубликована: Ноя. 1, 2024

Язык: Английский

Процитировано

4
Advanced machine learning techniques for predicting concrete mechanical properties: a comprehensive review of models and methodologies DOI
Fangyuan Li,

Md. Sohel Rana,

Muhammad Ahmed Qurashi

и другие.

Multiscale and Multidisciplinary Modeling Experiments and Design, Год журнала: 2024, Номер 8(1)

Опубликована: Дек. 18, 2024

Язык: Английский

Процитировано

3
Estimation of compressive strength of concrete with manufactured sand and natural sand using interpretable artificial intelligence DOI Creative Commons
Xiaodong Liu, Shengqi Mei, Xingju Wang

и другие.

Case Studies in Construction Materials, Год журнала: 2024, Номер 21, С. e03840 - e03840

Опубликована: Окт. 10, 2024

Язык: Английский

Процитировано

2
Interpretable machine‐learning models for predicting creep recovery of concrete DOI
Shengqi Mei, Xiaodong Liu, Xingju Wang

и другие.

Structural Concrete, Год журнала: 2024, Номер unknown

Опубликована: Окт. 16, 2024

Abstract Creep recovery of concrete is essential for accurately assessing the performance structures over service time. Existing creep models exhibit low accuracy, and influencing factors remain inadequately elucidated. In this paper, interpretable machine learning (ML) techniques were employed to develop a prediction model recovery. Several ML selected including random forest (RF), support vector regression (SVR), extreme gradient boosting (XGBoost) light (LGBM). order maximize sample size dataset, 109 sets data collected from existing literatures training. Feature selection utilized determine input parameters models, 12 variables selected. The fine‐tuned using Bayesian optimization techniques. To ensure reliability 10‐fold cross‐validation splitting implemented. results indicate that exhibited higher accuracy compared model. Among these LGBM demonstrated superior efficiency stability (with R 2 = 0.993, 0.978, 0.973 training, testing, validation sets, respectively). Shapley additive explanations (SHAP) interpret significance each parameter on prediction. Duration after unloading, stress magnitude, ambient relative humidity main feature Upon comparing factors, it was discerned there exists distinct difference between concrete.

Язык: Английский

Процитировано

0