Failure Sampling with Optimized Ensemble Approach for Structural Reliability Analysis of Complex Problems

Failure sampling is a structural reliability method based on modified conditional expectation suitable for complex problems which index–based approaches are inapplicable and simulation needed. Such tend to have nonsmooth limit-state boundaries or otherwise highly nonlinear. Previous studies recommended implementation of failure with an extrapolation technique using Johnson’s distribution the generalized lambda distribution. However, what works best problem-dependent. The uncertainty approach provides results particular problem limits potential effectiveness method. In this study, solution proposed issue that eliminates uncertainty. optimized ensemble forms uniquely weighted by utilizing predictive capability multiple curves maximize accuracy any problem. It was found produces solutions superior methods implementing previously presented in literature.