Data-Driven Approach: Predicting Critical Braking Temperature in Train Wheels Using Fiber Bragg Grating Sensors

This research presents a novel approach to predicting the critical temperature at which brakes must be applied to train wheels, combining real-time data collection through Fiber Bragg Grating (FBG) sensors with advanced machine learning models. While previous studies have either focused on theoretical models or utilized real-time sensor data without predictive capabilities, our approach uniquely integrates both sensor data and predictive modeling to improve accuracy. Conducted at The Oxford College of Engineering, our experiment used a scaled-down train model to measure key parameters such as speed, weight, and temperature during braking. Unlike traditional models, our study employed machine learning algorithms, including Linear Regression, Decision Tree, and Random Forest, to forecast the critical braking temperature. The Random Forest model significantly outperformed other models, achieving a perfect R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> score of 1.00 and a Mean Squared Error (MSE) of 0.00. Our integrated approach not only enhances the precision of braking temperature predictions but also offers a robust framework for predictive maintenance, improving the operational safety and efficiency of rail systems. This research marks a significant step forward in real-time monitoring and predictive analytics for rail transport systems, providing a superior method for detecting and responding to critical thermal conditions.