System reliability analysis by direct design point assessment

摘要

This paper presents a two steps computational strategy to evaluate the system probability of failure. First, the search of the design point using an unconstrained optimization algorithm. Second, the direct computation of the failure probability by any usual integration procedure, as the FORM/SORM analytical approximation, reduced integration, Monte Carlo importance sampling simulation, etc. This computer strategy can be accomplished, by defining a limit-state surface approximation to the envelope of the failure domain regions. The approximated function is updated in each step of the optimization process. Numerical applications are used to shown the robustness and accuracy of the proposed strategy.