Accéder directement au contenu Accéder directement à la navigation
Communication dans un congrès

Meta-model based optimisation algorithms for robust optimization of 3D forging sequences

Abstract : In order to handle costly and complex 3D metal forming optimization problems, we develop a new optimization algorithm that allows finding satisfactory solutions within less than 50 iterations (/function evaluation) in the presence of local extrema. It is based on the sequential approximation of the problem objective function by the Meshless Finite Difference Method (MFDM). This changing meta-model allows taking into account the gradient information, if available, or not. It can be easily extended to take into account uncertainties on the optimization parameters. This new algorithm is first evaluated on analytic functions, before being applied to a 3D forging benchmark, the preform tool shape optimization that allows minimizing the potential of fold formation during the two-stepped forging sequence.
Type de document :
Communication dans un congrès
Liste complète des métadonnées

https://hal-mines-paristech.archives-ouvertes.fr/hal-00510544
Contributeur : Magalie Prudon <>
Soumis le : jeudi 19 août 2010 - 14:58:39
Dernière modification le : mercredi 14 octobre 2020 - 04:02:44

Identifiants

Citation

Lionel Fourment. Meta-model based optimisation algorithms for robust optimization of 3D forging sequences. 10th ESAFORM Conference on Material Forming, Pts A and B, Apr 2007, Zaragoza, Spain. p. 21-26 - ISBN 0735404143, ⟨10.1063/1.2729482⟩. ⟨hal-00510544⟩

Partager

Métriques

Consultations de la notice

362