A Novel Graph-based Hierarchical Fault Propagation Modeling Approach for Complex Systems

摘要

Fault diagnosis in complex systems like radar is getting more and more difficult due to function interactions inside and among large numbers of functional units. Fault propagation modeling aims to model the fault propagation paths. It is of great significance for studying fault diagnostics. This work presents a graph-based hierarchical fault propagation modeling approach for complex systems. By reading and processing the model information filled in canonical forms, the system model is built automatically. Hierarchical modeling is realized by decomposing system model to line replaceable unit (LRU) models, which adopt fuzzy directed graph and achieve modeling of the redundant form of fault propagation. The method is employed for radar modeling and exhibit high flexibility and convenience of model management.