GEMS: A Python Library for Automation of Multidisciplinary Design Optimization Process Generation

François Gallard (1) , Charlie Vanaret (1) , Damien Guénot (1, 2) , Vincent Gachelin (1) , Rémi Lafage (3) , Benoit Pauwels (1) , Pierre-Jean Barjhoux (1) , Anne Gazaix (1)

1 IRT Saint Exupéry - Institut de Recherche Technologique
2 Atlantide (Altran technologies)
3 DTIS, ONERA, Université de Toulouse [Toulouse]

This paper describes the GEMS software developed as part of the IRT Saint Exupéry MDA-MDO project 1 for supporting Multidisciplinary Design Optimization (MDO) capabilities. GEMS is a Python library for programing MDO simulation processes, built on top of NumPy 2 , SciPy 3 and Matplotlib 4. GEMS aims at pushing forward the limits of automation in simulation process development , with a particular focus on : i) the automatic programming of MDO processes; ii) distributed and multi-level MDO formulations (or MDO architectures); iii) the integration and development of state-of-the-art algorithms for optimization, design of experiments, surrogate models and coupled analyses; iv) the automation of MDO result analysis; v) the deployment of MDO processes in heterogeneous and distributed industrial simulation environments.

Mots clés: OPTIMIZATION; ANALYTICAL TARGET; DESIGN

Domaines: Sciences de l'ingénieur [physics]; Physique [physics]; Mathématiques [math]; Informatique [cs]

François Gallard, Charlie Vanaret, Damien Guénot, Vincent Gachelin, Rémi Lafage, et al.. GEMS: A Python Library for Automation of Multidisciplinary Design Optimization Process Generation. AIAA/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, 2018, Jan 2018, KISSIMMEE, United States. ⟨10.2514/6.2018-0657⟩. ⟨hal-02335530⟩