TY  - INPR
A1  - Steiner, Maria
A1  - Bourinet, Jean-Marc
A1  - Lahmer, Tom
T1  - An adaptive sampling method for global sensitivity analysis based on least-squares support vector regression
N2  - In the field of engineering, surrogate models are commonly used for approximating the behavior of a physical phenomenon in order to reduce the computational costs. Generally, a surrogate model is created based on a set of training data, where a typical method for the statistical design is the Latin hypercube sampling (LHS). Even though a space filling distribution of the training data is reached, the sampling process takes no information on the underlying behavior of the physical phenomenon into account and new data cannot be sampled in the same distribution if the approximation quality is not sufficient. Therefore, in this study we present a novel adaptive sampling method based on a specific surrogate model, the least-squares support vector regresson. The adaptive sampling method generates training data based on the uncertainty in local prognosis capabilities of the surrogate model - areas of higher uncertainty require more sample data. The approach offers a cost efficient calculation due to the properties of the least-squares support vector regression. The opportunities of the adaptive sampling method are proven in comparison with the LHS on different analytical examples. Furthermore, the adaptive sampling method is applied to the calculation of global sensitivity values according to Sobol, where it shows faster convergence than the LHS method. With the applications in this paper it is shown that the presented adaptive sampling method improves the estimation of global sensitivity values, hence reducing the overall computational costs visibly.
KW  - Approximation
KW  - Sensitivitätsanalyse
KW  - Abtastung
KW  - Surrogate models
KW  - Least-squares support vector regression
KW  - Adaptive sampling method
KW  - Global sensitivity analysis
KW  - Sampling
Y1  - 2018
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:gbv:wim2-20181218-38320
N1  - This is the pre-peer reviewed version of the following article: https://www.sciencedirect.com/science/article/pii/S0951832017311808, which has been published in final form at https://doi.org/10.1016/j.ress.2018.11.015.
SP  - 1
EP  - 33
ER  - 
TY  - CHAP
A1  - Milbradt, Peter
A1  - Schwöppe, Axel
T1  - Finite Element Approximation auf der Basis geometrischer Zellen
N2  - Die Methode der Finiten Elemente ist ein numerisches Verfahren zur Interpolation vorgegebener Werte und zur numerischen Approximation von Lösungen stationärer oder instationärer partieller Differentialgleichungen bzw. Systemen partieller Differentialgleichungen. Grundlage dieser Verfahren ist die Formulierung geeigneter Finiter Elemente und Finiter Element Zerlegungen. Finite Elemente besitzen in der Regel eine geometrische Basis bestehend aus Strecken im eindimensionalen, Drei- oder Vierecken im zweidimensionalen und Tetra- oder Hexaedern im dreidimensionalen euklidischen Raum, eine Menge von Freiheitsgraden und eine Basis von Funktionen. Die geometrische Basis eines Finiten Elements wird verallgemeinert als geometrische Zelle formuliert. Diese geschlossene geometrische Formulierung führt zu einer geometrieunabhängigen Definition der Basisfunktionen eines Finiten Elements in den Zellkoordinaten der geometrischen Zelle. Finite Elemente auf der Basis geometrischer Zellen werden als Bestandteile Finiter Element Zerlegungen in Finiten Element Interpolationen und Finiten Element Approximationen verwendet. Die Finiten Element Approximationen werden am Beispiel der 2-dimensionalen Diffusionsgleichung über das Standard-Galerkin-Verfahren ermittelt.
KW  - Finite-Elemente-Methode
KW  - Approximation
Y1  - 2003
U6  - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:gbv:wim2-20111215-3333
ER  -