Coupling Sensitivity Analysis and Life Prediction of Multi-Degradation Variable Models

Based on the Copula function model, this paper proposes a method for predicting the equipment life under the coupling conditions of multiple degradation models. It considers the uncertainties of multiple degradation variables and degradation description models, extending the univariate and deterministic description models to the conditions of multivariate coupling and uncertain degradation description models. The degradation description model and Copula function are selected through degradation data hypothesis testing and Akaike Information Criterion (AIC), respectively. Sensitivity analysis is conducted on the impact of different degradation models on the multivariate j oint distribution under the same expectation conditions. The proposed life prediction method under multivariate coupling conditions is validated using a binary degradation variable simulation case.