文章基本信息

标题：Sensitivity analysis using stochastic expansion methods to calculate variance-based sensitivity indices
本地全文：下载
作者：V. Gregory Weirs ; V. Gregory Weirs ; Laura P. Swiler 等
期刊名称：Procedia - Social and Behavioral Sciences
印刷版ISSN：1877-0428
出版年度：2010
卷号：2
期号：6
页码：7765-7767
DOI：10.1016/j.sbspro.2010.05.218
语种：English
出版社：Elsevier
摘要：AbstractSensitivity analysis and uncertainty quantification are computationally expensive procedures. Stochastic expansion methods are an alternative approach for performing SA and UQ, and usually require fewer function evaluations. Recent work has extended non-intrusive stochastic expansion methods for calculating sensitivity indices. We examine the application of these techniques to a case in which data are generated from coupled, nonlinear partial differential equations. The values we analyze are generated from the numerical solution of the PDEs, in which we systematically vary both (i) fundamental modeling parameters and (ii) the underlying numerical algorithms. Our goal is to compare the performance of the sensitivity index calculations using sampling based methods and stochastic expansion methods, to gain an understanding of the strengths and weaknesses of both approaches.
关键词：Conservation laws;compressible flow;sensitivity analysis;uncertainty quantification;stochastic expansion