An application of dependent Kriging combined with spherical decomposition sampling for the system reliability analysis of flap mechanism

Reliability analysis for complex systems is a challenging problem, because of complex failure regions and frequently time-consuming simulations. Especially for complex systems with extremely rare events, it is of great significance to evaluate the reliability efficiently and accurately. Therefore, a novel reliability analysis method that combines the dependent Kriging method and adaptive spherical decomposition sampling for the rare event is proposed in this work to solve these problems. It makes full use of the mean and variance information of the Kriging predicted responses and the covariance information between the responses. In addition, the stopping criterion is directly related to the accuracy of failure probability rather than the accuracy of model construction. Furthermore, spherical decomposition sampling is used to estimate the small failure probability and improve sampling efficiency. Three test examples are presented to illustrate the accuracy and efficiency of the proposed method. Meanwhile, the flap motion mechanism is used as the research object to verify the application of the proposed method.