Dynamic Fault Trees Analysis for Importance Measures Based on Cut Sequence Set Model

Abstract

Importance Measures (IMs) for dynamic fault tree (DFT) remains widely open, for that Markov-based methodology is limited to exponential distributions as well as the state space explosion. Cut sequence describes the dynamic behaviors among components, and Cut sequence set (CSS) model provides ways to generate and calculate the probabilities of cut sequences, but it does not show how to analysis IMs using CSS model. This paper proposed a novel methodology of IMs in CSS model, including Birnbaum measure (BI), Risk Achievement Worth (RAW) and Structure Importance (SI) and an example demonstrates the superiority of the methodology. In contrast to the Markov-based methodology, the proposed methodology is combinatorial and can handle with any type of time-to-failure distributions for the system components.