An Information Geometry Approach to Robustness Analysis for the Uncertainty Quantification of Computer Codes

Robustness analysis is an emerging field in the uncertainty quantification domain. It involves analyzing response of a computer model—which has inputs whose exact values are unknown—to perturbation one or several its input distributions. Practical robustness methods therefore require coherent methodology for perturbing distributions; we present here such rigorous method, based on Fisher distance manifolds probability Further, provide numerical method to calculate perturbed densities practice which comes from Lagrangian mechanics and solving system ordinary differential equations. The introduced perturbations then used compute quantile-related indices. We illustrate these “perturbed-law based” indices models. also apply our industrial setting: simulation loss coolant accident nuclear reactor, where dozen model’s physical parameters not known exactly, limited knowledge their distributions available.