Investigating coupling patterns in state-space models for system reliability

State-space representation in system reliability modeling provides a flexible means for modeling dynamic scenarios, but monolithic models tends to be prohibitively complex for realistic systems. Breaking models into modules that can be analyzed independently (off-line) greatly reduces this complexity, but the characterization of the coupling among these modules should provide adequate accuracy once those modules are assembled together. In this paper, the feasibility of component-based modeling of shared repair resources is investigated. The work builds upon a recently developed compact representation of competing risks scenarios that matches the relative occurrence probability ratios of multiple risks (choice) and replaces it with a single Weibull distribution.