Adaptive RBF with hyperparameter optimisation for aeroacoustic applications

The present work reports an investigation on the use of adaptive metamodels based radial basis functions (RBFs) for aeroacoustic applications highly innovative configurations. relevance topic lies paramount importance metamodelling techniques within design optimisation process disruptive aircraft layouts. Indeed, air traffic growth, consequently hard environmental constraints imposed by regulations, will make a technological breakthrough, imperative need few years. As consequence, engineering community is paying particular attention to development unconventional For this class applications, designer cannot successfully rely historical data or low-fidelity models, and expensive direct simulations remain only valuable strategy. In regard, it can be demonstrated that surrogate i.e., metamodels, significantly reduces computing costs, especially in view robust approach optimised design. order further improve efficiency metamodel-based techniques, dynamic approaches hyperparameter sampling procedures have been recently developed. case study presented here pertains exploiting RBF-based noise shielding applications. analysis metamodel performances its convergence properties shows how final number reduced algorithm, still strongly depending choice RBF kernel function.