Approximation of reliability constraints by estimating quantile functions

A novel approach is proposed to effectively estimate the quantile functions of normalized performance indices of reliability constraints in a reliability-based optimization (RBO) problem. These quantile functions are not only estimated as functions of exceedance probabilities but also as functions of the design variables of the target RBO problem. Once these quantile functions are obtained, all reliability constraints in the target RBO problem can be transformed into non-probabilistic ordinary ones, and the RBO problem can be solved as if it is an ordinary optimization problem. Two numerical examples are investigated to verify the proposed novel approach. The results show that the approach may be capable of finding approximate solutions that are close to the actual solution of the target RBO problem.