Revolutionizing Fault Prediction in MetroPT Datasets: Enhanced Diagnosis and Efficient Failure Prediction through Innovative Data Refinement

Abstract

This scientific paper presents groundbreaking advancements in Predictive Maintenance (PdM) within Industry 4.0, employing cutting-edge machine learning classification algorithms for fault prediction and diagnosis in Air Production Unit (APU) systems like MetroPT and MetroPT -3. This research uses data-driven methodologies to optimize feature extraction techniques to enhance fault prediction and improve diagnostic accuracy. A robust and versatile model emerges through comprehensive testing, displaying exceptional potential in fault prediction and diagnosis for complex systems. The paper highlights the significance of enhanced analytical techniques, such as cross-validation, ensuring the reliability and robustness of the model, contributing to refined and accurate fault prediction and diagnosis, all without succumbing to overfitting. This work significantly advances the application of machine learning in predicting malignancy within Industry 4.0, showcasing the promise of these methodologies in fault prediction and diagnosis for intricate systems.