A BAYESIAN NETWORKS APPROACH FOR EVENT TREE TIME-DEPENDENCY ANALYSIS ON PHASED-MISSION SYSTEM

Event tree/fault tree (E/FT) method is the most recognized probabilistic risk assessment tool for complex large engineering systems, while its classical formalism most often only considers pivotal events (PEs) being independent or time-independent. However, the practical difficulty regarding phased-mission system (PMS) is that the PEs always modelled by fault trees (FTs) are explicit dependent caused by shared basic events, and phase-dependent when the time interval between PEs is not negligible. In this paper, we combine the Bayesian networks (BN) with the E/FT analysis to figure such types of PMS based on the conditional probability to give expression of the phase-dependency, and further expand it by the dynamic Bayesian networks (DBN) to cope with more complex time-dependency such as functional dependency and spares. Then, two detailed examples are used to demonstrate the application of the proposed approach in complex event tree time-dependency analysis.