Monte Carlo importance sampling optimization for system reliability applications

This paper focuses on the reliability analysis of multicomponent systems by the importance sampling technique, and, in particular, it tackles the optimization aspect. A methodology based on the minimization of the variance at the component level is proposed for the class of systems consisting of independent components. The claim is that, by means of such a methodology, the optimal biasing could be achieved without resorting to the typical approach by trials.