Dr. Ing. Laura Mainini Raytheon Technologies Engineering problems are characterized by a growing complexity that motivates the need to both manage interactions among multi-physics systems and contain analysis and development cost. While complexity is often mentioned as a challenge, it can also be an opportunity to improve system design and performance. Many strategies, algorithms and architectures have been proposed in the last decades and are commonly referred to as Multidisciplinary Design Optimization (MDO) methodologies. MDO largely grew in the domain of aerospace sciences, but currently comprises approaches to address the optimal design of a diverse spectrum of multidisciplinary and multi-component engineering systems. Another active area of research regards strategies to reduce computational burden and analysis cost, in particular through variable fidelity methods. Multifidelity approaches combine analysis models characterized by different levels of accuracy and may include approximations obtained with reduced order modeling and surrogate modeling techniques. The course aims to provide the students with the tools to familiarize with methods and computational strategies for digital engineering that can be applied to a variety of engineering problems, from multidisciplinary design analysis and optimization to multi-source information handling for time and resource constrained computational tasks.

-

The course is organized into modules as follows: • Course Introduction: motivation and rationale; learning objectives, course organization, team activities and final assessment criteria. • Module-1 Introduction to MDO: Goal, role, formulation, framework, constitutive elements, main challenges. | MDO and Digital Engineering • Module-2 Search methods and optimization procedures: optimality conditions, unconstrained problems, constrained problems | Multimodal optimization, multi-objective and multicriteria optimization strategies. • Module-3 Decomposition: criteria and assessment | Multidisciplinary Analysis problem formulation and evaluation schemes | MDO Architectures: single-level monolithic architectures, multiple-level distributed architectures. • Module-4 Approximation method for in MDO: role and use; approximation methods: classification and characteristics | Multifidelity methods for approximation, optimization and information fusion.

A distanza in modalità sincrona

Presentazione report scritto

P.D.2-2 - Maggio