In this work, we explore how interval-based approaches can be used to obtain a solution to the inverse problem, in particular when measurements are inconsistent with one another. We applied this strategy to identify the elastic parameters of an isotropic material, and it allows us to take into account both prior information about the parameters and measurement uncertainty in the form of sets (intervals or boxes) during the inversion process. The main advantage of this strategy is not only that it helps to obtain a feasible set of parameters, but that it is also able to detect the outliers among noisy measurements. This particular chapter explores how this strategy can be used in combination to surrogate modeling for identification problems in high dimensional settings.
Dealing with High Dimensional Inconsistent Measurements in Inverse Problems Using Surrogate Modeling: An Approach Based on Sets and Intervals
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 ; Kapitel : 26 ; 421-433
2021-07-16
13 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Inverse Surrogate Modeling: Output Performance Space Sampling
British Library Conference Proceedings | 2010
|