Review on Wheelset Maintenance Strategies Using Physical and Data-Driven Models in Railway Transportation

This review aims to investigate the effectiveness of using physical-driven, data-driven models and their combination models for wheelset maintenance strategies in railway transportation. The study examines various physical and data-driven models and combination model to discuss the combination maintenance approach that can improve the reliability and safety of railway systems. The advantages and limitations of typical wheelset maintenance strategies, physical model and data-driven model are also discussed, including their ability to accurately predict wheelset faults and reduce maintenance costs. The review of the literature suggests that the combination of physical and data-driven models can lead to more effective and efficient maintenance strategies. However, challenges such as data quality, model accuracy and computational complexity must be considered when implementing/designing strategies. The study summarises that the wheelset maintenance strategy based on the combination of physical and data-driven models is a promising approach for improving the maintenance of railway wheelsets and could potentially lead to more reliable and safe railway transportation.