Failure propagation analysis of high-speed train systems from the perspective of multi-layer stochastic flow network

The failure of minimum maintenance units (MMUs) in a high-speed train system (HSTS) increases the workloads of other MMUs, which can lead to failure propagation (FAP) and cause catastrophic accidents. In this study, a model is developed to analyze the process of FAP in HSTSs, considering system topology, dynamic functions, failure modes (FMs) and degradation of MMUs. Firstly, the entire topology and system functions of the HSTS are constructed as a multi-layer stochastic flow network (MSFN), and a network-state equation is established to analyze the flows along the edges. Subsequently, the initial load is defined based on the flows of the edges and load influencing rate (LIR), which quantifies the degree of initial load loss for an MMU under specified FMs at the initial failure time. Finally, a load redistribution model is used to introduce the inter-layer propagation mechanism, and different load redistribution strategies are adopted within different layers to predict fault propagation paths, providing theoretical support for maintenance decision-making. A practical case from the “Fuxing†HSTS is presented to illustrate the analysis procedure and compare the effectiveness of the proposed model with those of the existing methods.