The increasing constraints on the design of electromagnetic devices require numerical tools that are able to finely model the electromagnetic fields in the studied domain. Finite Element Method (FEM) is the most used tool to satisfy such a requirement. However, it may turn out that this tool is very expensive in computational time due to nonlinear behavior, 3D geometries, and time dependency. Thus, its usage for optimization, i.e. iterative process, should be made with caution since only a limited number of evaluations of the simulation tool are possible. Thus, the use of some specific algorithms may not be possible in a limited time, e.g. genetic algorithms, that require many evaluations of the FEM code. On the other hand, gradient-based algorithms cannot be used to their full potential due to the re-meshing error that may appear when computing the gradient using finite difference. Le Jeudi 21 mai 2018, Reda EL Bechari,doctorant de l’\u00e9quipe OMN,\u00a0effectuera un s\u00e9minaire sur le th\u00e8me : Approaches to Design Optimization of Electromagnetic Devices using Finite Element Method Date :\u00a0Jeudi 21 juin 2018 \u00e0 13h30 Lieu : Salle 2S53 \u00e0 Lilliad The increasing constraints on the design of electromagnetic devices require numerical tools that are […]<\/p>\n","protected":false},"author":8,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":[],"categories":[34,37],"tags":[],"_links":{"self":[{"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/posts\/5047"}],"collection":[{"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/comments?post=5047"}],"version-history":[{"count":2,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/posts\/5047\/revisions"}],"predecessor-version":[{"id":5049,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/posts\/5047\/revisions\/5049"}],"wp:attachment":[{"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/media?parent=5047"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/categories?post=5047"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/l2ep.univ-lille.fr\/en\/wp-json\/wp\/v2\/tags?post=5047"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
\nThere exist two approaches. A non-intrusive approach that considers the FEM simulation as a black-box and constructs cheap meta-models to reduce the computational burden while refining only in promising regions. And a second approach aims to exploit the derivative of quantities of interest computed from the finite element code by using the adjoint method to be able to use gradient-based algorithms.<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"