Search

Copula model has been increasingly used in reliability modeling of correlation systems. In this paper, the static copula model, time-varying copula model and variable structure copula model are compared. Firstly, variable structure copula model is introduced. Secondly, the procedures of Bayes diagnosis method which is used to diagnose the variable structure points are introduced. Finally, the correctness of Bayes diagnosis method for finding variable structure points is verified by a simple numerical example. The comparison results illustrate that the staged correlation parameter estimated by variable structure Copula model is better than the constant correlation parameter estimated by static Copula model, and also it has a consistent trend with correlation parameter estimated by time-varying Copula model.