System-Level Fault Prognosis for High-Speed Railway On-Board Systems

The on-board system is the core part of a train control system that ensures traffic safety and improves operational efficiency. At present, fault data processing methods for on-board systems do not show the degradation of components, subsystems, or the system over time, but only focus on identifying faults after a failure has occurred, so fault prognosis cannot be realized. This paper applies a multilevel flow model based on system-level failure mode and effect analysis. The model defines the primary and secondary alarms, divides the failure modes into different levels, and ranks the risk priority numbers at different levels. The Balise transmission system is then used as an example to demonstrate the system-level fault prognosis methodology, and the variables from different levels are combined through the component model, subsystem model, and system model. The results are verified according to the actual operation manual.