A Topology-Based Method for Quantifying and Evaluating the System Structural Fault Tolerance Capability

In modern complex systems, the capability to sustain operations despite component failures is defined as the fault tolerance capability. To ensure system reliability, it is essential to quantify and evaluate the fault tolerance capability systematically. However, existing research primarily focuses on failure propagation and lacks the consideration of critical thresholds for fault tolerance, leading to inadequate guidance for fault-tolerant system. To tackle these challenges, this paper proposes a topology-based method for quantifying and evaluating the system structural fault tolerance. By modelling systems as the directed graph based on the functional logic, we introduce the all functional paths (AFP) to measure redundancy through valid input-to-output paths. The structural fault tolerance index is derived by normalizing AFP values under fault conditions against a fault-free baseline, which is applicable to both single-input single-output (SISO) and multi-input multi-output (MIMO) systems. Additionally, we define the fault tolerance threshold based on the minimum redundancy requirement and introduce the margin to assess compliance. A case study on the redundant flywheel system (RFS) demonstrates the method's feasibility, and we calculate the structural fault tolerance index and the corresponding fault tolerance margin of the RFS in each state. The proposed framework bridges component-level faults to system-level functionality, offering actionable insights for evaluating and improving redundant system.

Keywords: Fault tolerance, Structural diagram model, Margin, Redundant flywheel system.