Investigation of machine learning models in predicting compressive strength for ultra-high-performance geopolymer concrete: A comparative study

Construction and Building Materials, Год журнала: 2024, Номер 436, С. 136884 - 136884

Опубликована: Июнь 12, 2024

Язык: Английский

---

Modelling the compressive strength of geopolymer recycled aggregate concrete using ensemble machine learning

Advances in Engineering Software, Год журнала: 2024, Номер 191, С. 103611 - 103611

Опубликована: Март 1, 2024

In the quest to reduce environmental impact of construction sector, adoption sustainable and eco-friendly materials is imperative. Geopolymer recycled aggregate concrete (GRAC) emerges as a promising solution by substituting supplementary cementitious materials, including fly ash slag cement, for ordinary Portland cement utilizing aggregates from demolition waste, thus significantly lowering carbon emissions resource consumption. Despite its potential, widespread implementation GRAC has been hindered lack an effective mix design methodology. This study seeks bridge this gap through novel machine learning (ML)-based approach accurately model compressive strength (CS) GRAC, critical parameter ensuring structural integrity safety. By compiling comprehensive database existing literature enhancing it with synthetic data generated tabular generative adversarial network, research employs eight ensemble ML techniques, comprising three bagging five boosting methods, predict CS high precision. The models, notably extreme gradient boosting, light categorical regressors, demonstrated superior performance, achieving mean absolute percentage error less than 6 %. precision in prediction underscores viability optimizing formulations enhanced applications. identification testing age, natural fine content, ratio pivotal factors offers valuable insights into process, facilitating more informed decisions material selection proportioning. Moreover, development user-friendly graphical interface exemplifies practical application research, potentially accelerating mainstream practices. enabling use contributes global effort promote within industry.

Язык: Английский

---

A systematic literature review of AI-based prediction methods for self-compacting, geopolymer, and other eco-friendly concrete types: Advancing sustainable concrete

Construction and Building Materials, Год журнала: 2024, Номер 440, С. 137370 - 137370

Опубликована: Июль 16, 2024

Язык: Английский

---

Multi-performance optimization of low-carbon geopolymer considering mechanical, cost, and CO2 emission based on experiment and interpretable learning

Construction and Building Materials, Год журнала: 2024, Номер 425, С. 136013 - 136013

Опубликована: Апрель 1, 2024

Язык: Английский

---

Effect of Data Augmentation Using Deep Learning on Predictive Models for Geopolymer Compressive Strength

Applied Sciences, Год журнала: 2024, Номер 14(9), С. 3601 - 3601

Опубликована: Апрель 24, 2024

In recent years, machine learning models have become a potential approach in accurately predicting the concrete compressive strength, which is essential for real-world application of geopolymer concrete. However, precursor system known to be more heterogeneous compared Ordinary Portland Cement (OPC) concrete, adversely affecting data generated and performance models. To its advantage, enrichment through deep can effectively enhance prediction Therefore, this study investigates capability tabular generative adversarial networks (TGANs) generate on mixtures strength It assesses impact using synthetic with various models, including tree-based, support vector machines, neural networks. For purpose, 930 instances 11 variables were collected from open literature. particular, 10 content fly ash, slag, sodium silicate, hydroxide, superplasticizer, fine aggregate, coarse added water, curing temperature, specimen age are considered as inputs, while output A TGAN was employed an additional 1000 points based original dataset training new predictive These evaluated real test sets trained data. The results indicate that developed significantly improve performance, particularly networks, followed by tree-based machines. Moreover, characteristics greatly influence model both before after augmentation.

Язык: Английский

---

Research on seismic performance prediction of CFST latticed column-composite box girder joint based on machine learning

Construction and Building Materials, Год журнала: 2025, Номер 460, С. 139811 - 139811

Опубликована: Янв. 1, 2025

Язык: Английский

---

Development and optimization of geopolymer concrete with compressive strength prediction using particle swarm-optimized extreme gradient boosting

Applied Soft Computing, Год журнала: 2025, Номер unknown, С. 113149 - 113149

Опубликована: Апрель 1, 2025

Язык: Английский

---

Using ensemble machine learning and metaheuristic optimization for modelling the elastic modulus of geopolymer concrete

Cleaner Materials, Год журнала: 2024, Номер 13, С. 100258 - 100258

Опубликована: Июнь 27, 2024

Geopolymer concrete emerges as a sustainable and durable alternative to conventional concrete, addressing its high carbon footprint enhanced durability. The distinct properties of geopolymer governed by supplementary cementitious materials alkaline activators, promise reduced environmental impact improved structural resilience. However, complex composition complicates the prediction mechanical such elastic modulus, crucial for applications. This study introduces an innovative approach using eXtreme Gradient Boosting (XGBoost) technique integrated with multi-objective grey wolf optimizer model modulus concrete. By dynamically selecting influential features optimizing accuracy, this methodology advances beyond traditional empirical models, which fail capture nonlinear interactions intrinsic Utilizing comprehensive database gathered from extensive literature, 22 potential variables were examined that influence concrete's modulus. After mitigating multicollinearity hyperparameters via Bayesian optimization, six XGBoost models developed different combinations input variables, revealing compressive strength total water content pivotal predictors. findings illustrate models' precision, trade-off between accuracy simplicity visualized through relationship number error. culminates in user-friendly graphical user interface enables easy fosters educational engagement. interface, available online, underscores practicality accessibility advanced machine learning predictions. Overall, research not only provides robust predictive framework optimized but also enhances understanding underlying determinants, contributing advancement construction materials.

Язык: Английский

---

Ensemble machine learning models for predicting the CO2 footprint of GGBFS-based geopolymer concrete

Journal of Cleaner Production, Год журнала: 2024, Номер 472, С. 143463 - 143463

Опубликована: Авг. 26, 2024

While geopolymer concrete (GPC) has gained popularity for its environmentally friendly attributes compared to ordinary Portland cement, the absence of a prediction model carbon footprint constituents presents challenges optimization within evolving industry.This study offers thorough CO 2 ground granulated blast-furnace slag (GGBFS)-based GPC, utilizing advanced AI techniques, including combination machine learning models and stacking ensembles.This research statistically examines crucial parameters responsible emissions in GGBFS-based GPC production, identifying factors like superplasticizer content, initial curing temperature, NaOH (dry) content as significant contributors.Emphasizing sustainability, advocates optimizing mixtures by considering ratio other activator materials.After rigorously evaluating 12 models, ensemble this identified M4-a Support Vector Regression (SVR) Neural Network (NN)-as weak Decision Tree (DT) meta-model, most effective predicting footprints.The choice M4 is supported various performance metrics such lowest Mean Squared Error 88.8 Root 9.42, alongside highest R , Adjusted Explained Variance scores, all approximately 0.95.Additional analyses, Euclidean distance Taylor diagrams, further substantiate selection M4.The findings have practical implications sustainable cleaner enabling businesses optimize GPC.

Язык: Английский

---

Deep learning–based prediction of compressive strength of eco-friendly geopolymer concrete

Environmental Science and Pollution Research, Год журнала: 2024, Номер 31(28), С. 41246 - 41266

Опубликована: Июнь 7, 2024

The greenhouse gases cause global warming on Earth. cement production industry is one of the largest sectors producing gases. geopolymer produced with synthesized by reaction an alkaline solution and waste materials such as slag fly ash. use eco-friendly concrete decreases energy consumption In this study, f

Язык: Английский

---