Development of Data-Driven Methods for Railway In-Train Force Monitoring and Damage-Tolerant Maintenance for Railway Couplings

In modern trains, large dynamic forces act on railway couplings, mechanical connections between carriages, leading to fatigue issues over time. This research helps improve safety and reduce maintenance costs by developing data-driven methods to monitor and maintain these couplings. A machine learning model is developed to estimate forces that they transmit during operation without the need for costly sensors. The study also introduces an advanced simulation approach to predict how fatigue cracks grow in coupling parts during real-world service. Together, these innovations support a shift from fixed schedules to smarter, condition-based maintenance, improving reliability and extending the life of components.