Surrogate-Based Optimization of System Architectures Subject to Hidden Constraints - Archive ouverte HAL
Communication Dans Un Congrès Année : 2024

Surrogate-Based Optimization of System Architectures Subject to Hidden Constraints

Résumé

System architecture optimization can support the design of novel architectures by formulating the architecting process as an optimization problem. The exploration of novel architectures requires physics-based simulation due to a lack of prior experience to start from, which introduces two specific challenges for optimization algorithms: evaluations become more expensive (in time) and evaluations might fail. The former challenge is addressed by Surrogate-Based Optimization (SBO) algorithms, in particular Bayesian Optimization (BO) using Gaussian Process (GP) models (Kriging). An overview is provided of how BO can deal with challenges specific to architecture optimization, such as design variable hierarchy and multiple objectives: specific measures include ensemble infills, a multi-step infill optimization, and hierarchical sampling and correction algorithms. Evaluations might fail due to non-convergence of underlying solvers or infeasible geometry in certain areas of the design space. Such failed evaluations, also known as hidden constraints, pose a particular challenge to SBO/BO, as the surrogate model cannot be trained on empty results. This work investigates various strategies for satisfying hidden constraints in BO algorithms. Three high-level strategies are identified: rejection of failed points from the training set, replacing failed points based on non-failed points, and predicting the failed region. Through investigations on a set of test problems including a jet engine architecture optimization problem, it is shown that best performance is achieved with a mixed-discrete GP to predict the Probability of Viability (PoV), and by ensuring selected infill points satisfy some minimum PoV threshold. The jet engine architecture problem features a 50% failure rate and could not previously be solved by a BO algorithm. The developed BO algorithm and used test problems are available in the open-source Python library SBArchOpt.
Fichier principal
Vignette du fichier
DTIS2024-018-DTIS2024-018 posprint-Acceptée.pdf (2.18 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04462408 , version 1 (16-02-2024)
hal-04462408 , version 2 (21-08-2024)

Identifiants

Citer

Jasper H Bussemaker, Paul Saves, Nathalie Bartoli, Thierry Lefebvre, Björn Nagel. Surrogate-Based Optimization of System Architectures Subject to Hidden Constraints. AIAA Aviation Forum and ASCEND co-located Conference Proceedings, Jul 2024, Las Vegas, United States. ⟨10.2514/6.2024-4401⟩. ⟨hal-04462408v2⟩

Collections

ONERA DTIS_ONERA
114 Consultations
130 Téléchargements

Altmetric

Partager

More