Split Tensile Strength Prediction of Recycled Aggregate-Based Sustainable Concrete Using Artificial Intelligence...

Area of Science:

• Civil Engineering
• Materials Science
• Artificial Intelligence

Background:

• Sustainable concrete development is crucial for environmental protection and material demand.
• Recycled aggregate (RA) is a key component in eco-friendly concrete formulations.
• Accurate prediction of concrete properties is essential for its practical application.

Purpose of the Study:

• To predict the split tensile strength (STS) of concrete incorporating recycled aggregate (RA) using artificial intelligence (AI).
• To compare the performance of artificial neural network (ANN), decision tree (DT), and random forest (RF) models for this prediction task.

Main Methods:

• Utilized three machine learning techniques: ANN, DT, and RF.
• Employed statistical tests and k-fold cross-validation for model validation.
• Conducted sensitivity and SHAP analyses to determine input parameter importance.

Main Results:

• The random forest (RF) model exhibited the highest precision in predicting the STS of RA-based concrete.
• RF model performance was validated by a high coefficient of determination and low error metrics (MAE, MSE, RMSE).
• Sensitivity and SHAP analyses provided insights into the influence of input parameters.

Conclusions:

• The RF model is highly accurate and precise for predicting the STS of sustainable concrete containing RA.
• AI, particularly RF, offers a powerful tool for optimizing the design of eco-friendly concrete.
• Understanding parameter importance enhances the reliability and interpretability of AI models in construction materials research.