This work aims to explore the potential of Gaussian processes to reduce the computational load of robust optimisation problems faced with the risk function methodology. In particular, the focus is on the conditional value-at-risk function (CVaR) and shows how a reduced number of samples can be used to obtain an approximation of CVaR usable in a robust optimisation loop based on evolutionary algorithms. The method shows its effectiveness through the application to the robust aerodynamic shape design of an airfoil in the transonic regime.
Gaussian Processes for CVaR Approximation in Robust Aerodynamic Shape Design
Space Technol.Proceedings
International Conference on Uncertainty Quantification & Optimisation ; 2020 ; Brussels, Belgium November 17, 2020 - November 20, 2020
Advances in Uncertainty Quantification and Optimization Under Uncertainty with Aerospace Applications ; Chapter : 20 ; 327-346
2021-07-16
20 pages
Article/Chapter (Book)
Electronic Resource
English
Risk Measures Applied to Robust Aerodynamic Shape Design Optimization
Springer Verlag | 2020
|