An Ensemble Learning Approach for Predicting Compressive Strength of Recycled Aggregate Concrete

Abstract

The construction industry is one of the largest consumers of natural resources and a major contributor to environmental degradation. With increasing emphasis on sustainable development, the use of recycled concrete aggregate (RCA) has gained significant attention. However, predicting the compressive strength of recycled concrete remains a challenging task due to the variability in material properties. This project proposes an ensemble learning-based model to accurately predict the compressive strength of recycled concrete using multiple input parameters.The system utilizes machine learning techniques to analyze the relationship between various factors such as cement content, water ratio, recycled aggregate proportion, fine and coarse aggregates, fly ash, superplasticizer, and curing age. These parameters significantly influence the strength characteristics of concrete. Traditional empirical methods often fail to capture the nonlinear relationships between these variables, leading to inaccurate predictions.The proposed approach employs an ensemble learning model, which combines multiple machine learning algorithms to improve prediction accuracy and robustness. The model is trained using a dataset containing various concrete mix designs and their corresponding compressive strength values. By leveraging ensemble techniques, the system reduces overfitting and enhances generalization performance.A web-based application is developed using the Django framework to provide an interactive interface for users. The application allows users to input material parameters and obtain predicted compressive strength values instantly. The backend integrates the trained machine learning model using Joblib, enabling efficient loading and prediction. The results demonstrate that the ensemble learning model outperforms traditional regression models in terms of accuracy and reliability. The system provides a practical tool for engineers and researchers to estimate concrete strength without conducting extensive laboratory experiments, thereby saving time and cost.This project contributes to sustainable construction practices by promoting the use of recycled materials and providing a reliable method for strength prediction. The proposed system can be extended with additional features such as real-time data integration, advanced visualization, and optimization techniques for mix design