The course introduces the theoretical and practical foundations of multidisciplinary optimization (MDO) and reduced-order models in the aerospace context. It will address constrained multi-objective problems, surrogate modeling techniques, and the integration of models with different levels of fidelity, with particular attention to uncertainty management and design robustness. The practical sessions will enable participants to develop optimization workflows and apply advanced methodologies using industrial tools, with reference to realistic case studies in the field of aerospace propulsion.

Basic knowledge of physics and mathematics

Introductory Module on MDO Optimization in the Aerospace Field (4 hours) · Formulation of constrained multi-objective problems · Exploration of the parameter space · Dimensionality reduction of the parameter space · Multifidelity approaches · Examples of aerospace applications Reduced Order Modeling Module (4 hours) · Surrogate models and response surfaces · Proper Orthogonal Decomposition · Nonlinear registration-based models · Machine learning and explainable AI · Application case: ROM generation using RomBOX Shape Optimization Module (4 hours) · Comparison of optimization algorithms: exploration vs. exploitation · Integration of an optimization process in an industrial context · Decision-making strategies in multi-objective problems: sensitivity analysis and Pareto front · Application case: multidisciplinary optimization using modeFRONTIER

In presenza

Presentazione orale

P.D.2-2 - Ottobre