APPLICATION OF ARTIFICIAL INTELLIGENCE IN THE ANALYSIS OF TURNING PARAMETERS IN THE PRODUCTION AND MAINTENANCE OF RAILWAY COMPONENTS

Abstract

Maintenance and manufacturing of railway components have a direct impact on the safety and operational efficiency of the railway system. Turning, as a finishing process applied to wheels and rails, significantly influences wear behavior, vibration levels, and noise generation. Conventional approaches to optimizing turning parameters are largely based on operator experience and empirical methods, which may result in variations in quality and an increased risk of in-service failures. This study investigates the application of artificial intelligence (AI) in the analysis and optimization of turning parameters for monoblock wheels made of alloy steel. Machine learning models—artificial neural networks (ANN), Random Forest, and Support Vector Machines (SVM)—were employed to predict wear progression, identify influential process parameters, and classify surface quality. The results indicate that AI-based models enable accurate prediction of wear and deformation, identification of optimal turning regimes, and adaptive process control in real time. The implementation of these algorithms contributes to extended component service life, reduced operating costs, and improved overall railway system safety.