An implicit method for probabilistic common-cause failure analysis using Bayesian Network

摘要

A probabilistic common-cause failure (PCCF) is any condition or event that causes multiple components fail or malfunction simultaneously with different occurrence probabilities. A system subject to PCCFs is sometimes affected by multiple common causes (CCs) whose relationship may be very complex. This paper proposes an implicit method to model systems subject to PCCFs using Bayesian Network (BN). Three kinds of relationships between CCs are considered: s-independence, s-dependence or mutually exclusion. The proposed method has no limitation on the type of failure distributions of system components. Finally, the proposed method is illustrated by an example computer system.