A Reliability Analysis Method for an Aircraft Power System with Hybrid Uncertainty and Failure Dependence

Abstract

Because of the increasing complexity of modern electromechanical systems, it is important to develop a reliability analysis model for systems with hybrid uncertainty and failure dependence. In practice, engineers may not have enough information to estimate the correlation between components/subsystems, which is named as non-deterministic dependence problem. In order to model the impact of uncertainty and non-deterministic dependence on system performance, this paper proposed a reliability analysis method based on Bayesian network and affine arithmetic. The proposed method can be used to overcome the limitation of interval arithmetic and retains the advantages of Bayesian network in dealing with the problems of correlation and uncertainty. Finally, the proposed affine-based Bayesian network is applied to an aircraft power system.