Application of AI Techniques for Asphalt Concrete Mix Production Optimization

Abstract

In present-day conditions of road infrastructure development, ensuring the high quality of asphalt concrete mixes contributes to the durability and reliability of road pavements. This article investigates the application of artificial intelligence techniques to analyze asphalt quality aimed at optimizing production and improving the reliability of road pavements. This study introduces a pioneering approach to asphalt concrete mix quality enhancement using artificial intelligence (AI) techniques, specifically artificial neural networks (ANN) and least-squares support vector machine (LS-SVM). The application of these methods allows for carrying out efficient analysis of data, reflecting asphalt quality, predicting asphalt characteristics, and optimizing production processes. The authors conducted experiments using real asphalt properties, which were used to train and set ANN and LS-SVM models. The obtained results were compared with existing methods of asphalt quality analysis. The conducted analysis confirmed the effectiveness of using ANN and SVM to analyze asphalt quality. This approach provides an opportunity for accurate prediction of asphalt performance characteristics and production process optimization, contributing to the improvement of the durability and reliability of road pavements. The obtained results have practical significance for engineers and specialists in the field of road infrastructure construction and maintenance. The results of the study validate the superiority of AI-driven models in achieving precise and reliable asphalt mix designs, marking a considerable advancement over traditional methods.