Toward Objective Assessment of Simulation Predictive Capability

Two different metrics quantifying model and simulator predictive capability are formulated evaluated; both exploit results from conducted validation experiments where simulation compared to the corresponding measured quantities. The first metric is inspired by modified nearest neighbor coverage second Kullback–Liebler divergence. two implemented in Python a here-developed general metamodel designed be applicable for most physics-based models. These implementations together facilitate offline online evaluation. Additionally, connection between two, here separated, concepts of credibility established realized metamodel. are, finally, evaluated an aeronautical domain context.