THE RELIABILITY INDEX APPROACH WITH EVOLUTIONARY ALGORITHMS: APPLICATION TO THE RBRDO PROBLEM OF COMPOSITE STRUCTURES

Abstract

A new methodology to solve the Reliability Index Approach (RIA) minimization problem with high levels of accuracy and still with practical efficiency is presented. This method is based on the exclusive use of genetic algorithms with elitist strategy, to make possible to find the global Most Probable Failure Point, in the uncertainty space. To overcome the computational costs inherent to these class of algorithms, the RIA minimization problem is reformulated and a new heuristic is proposed. The method is then applied to the Reliability based Robust Design Optimization problem of composite structures, where minimum weight and maximum system robustness are objectives and the target reliability index is a constraint.