Review on Research on Dependent Failure Analysis of Complex Systems

With systems developing towards complication, structuration and hierarchicalization, the deficiencies of traditional reliability assessment and analysis on the basis of independence and binary states become serious gradually. The status and impact of dependent failure analysis become more significant in reliability assessment and analysis of complex systems with redundant components. On the basis of introducing dependent failure, common cause failure is defined and applied to expand traditional fault tree. Throughout the development of dependent failure analysis, the advantages and disadvantages of five common methods concerned are summarized, including beta factor model, basic parameter model, multiple Greek letter, alpha factor model and binomial failure rate model. And the research directions of dependent failure analysis of complex systems in the future are predicted, i.e., dependent failure analysis under uncertain conditions and dependent failure analysis for multi-states systems.