Abstract: In order to solve the uncertainty of basic events’ failure possibilities in T-S fault tree due to poor fault data and complex fault mechanisms, and overcome applicability shortcomings of probability and fuzzy T-S fault tree and Boolean logic gate fault tree, convex model T-S fault tree and importance analysis methods are proposed: Interval T-S fault tree is proposed, which combines interval model with T-S fault tree and utilizes interval model to describe the failure probabilities of basic events as not easy to obtain accurately; On this basis, hyper-ellipsoid T-S fault tree is proposed, which utilizes hyper-ellipsoid model to define the ranges of the uncertain parameters to improve the relatively conservative reliability indices calculated by the interval T-S fault tree method; Then the importance indices of convex model T-S fault tree are presented to provide basis for finding the key link of the system; The failure possibilities of super events and T-S convex model importance and T-S convex model criticality importance of basic events are obtained, and the feasibility of the proposed methods are proved by comparing with convex model Boolean logic gate fault tree and T-S fault tree methods. At last, the proposed convex model T-S fault tree and importance analysis methods are applied to reliability analysis of an integrated navigation system.

Key words: convex model, importance, reliability analysis, T-S fault tree