We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Prediction of moisture susceptibility of asphalt mixtures containing RAP materials using machine learning algorithms.
- Authors
Afshin, Abolfazl; Behnood, Ali
- Abstract
This study aims to investigate the application of Machine Learning (ML) methods for predicting the moisture susceptibility of asphalt mixtures containing Reclaimed Asphalt Pavement (RAP) materials. Four ML algorithms were utilised for this purpose, including Decision Tree (M5P), Random Forest (RF), Gradient Boosting (GB) and Extreme Gradient Boosting (XGBoost). The most influential variables in the moisture performance of asphalt mixtures were identified using a comprehensive database comprising mixture properties, binder properties, recycled materials and environmental conditions. The study underscores the potential of ML models in improving the reliability and efficiency of moisture susceptibility predictions, ultimately contributing to the development of more durable and resilient asphalt pavements. The results of this study demonstrated that ML models possess a strong capability in estimating Tensile Strength Ratio (TSR) values, signifying a pivotal step in developing machine-based predictive models for assessing and improving the durability of asphalt pavements. Notably, the XGBoost model exhibited the highest accuracy with 0.96 (training) and 0.75 (testing) R2 values. Comparison of predicted and actual TSR values and residual analysis were explored to assess ML models robustness and validity, demonstrating the potential of ML models in improving the reliability and efficiency of moisture-induced damage predictions.
- Subjects
ASPHALT pavement recycling; ASPHALT pavements; DATABASES; DECISION trees; RANDOM forest algorithms
- Publication
International Journal of Pavement Engineering, 2024, Vol 25, Issue 1, p1
- ISSN
1029-8436
- Publication type
Academic Journal
- DOI
10.1080/10298436.2024.2431610