Active Kriging-Based Adaptive Importance Sampling for Reliability and燬ensitivity Analyses of Stator Blade Regulator

The reliability and sensitivity analyses of stator blade regulator usually involve complex characteristics like high-nonlinearity, multi-failure regions, small failure probability, which brings in unacceptable computing efficiency accuracy the current analysis methods. In this case, by fitting implicit limit state function (LSF) with active Kriging (AK) model reducing candidate sample pool adaptive importance sampling (AIS), a novel AK-AIS method is proposed. Herein, AK Markov chain Monte Carlo (MCMC) are first established to identify most probable region(s) (MPFRs), kernel density estimation (AKDE) constructed select samples. With best samples sequentially attained reduced employed update Kriging-fitted LSF, probability indices acquired at lower cost. proposed verified two numerical examples, then applied typical regulator. methods comparison, proven hold advantages on problems.