Reliability-oriented sensitivity analysis in presence of data-driven epistemic uncertainty

Reliability assessment in presence of epistemic uncertainty leads to consider the failure probability as a quantity depending on state knowledge about uncertain input parameters. The joint distribution is often learnt from small-sized dataset provided by operating experience. computed depends estimated marginal distributions and copula distribution. This paper develops reliability-oriented sensitivity analysis procedure order measure influence exerted data-driven modeling both margins copula. proposed methodology validated for deterministic stochastic reliability methods through an extensive simulation study including several analytical performance functions well real-life code dealing with buckling laminated composite plate.