Distribution-Based Global Sensitivity Analysis By Polynomial Chaos Expansion

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26110%2F20%3APU138598" target="_blank" >RIV/00216305:26110/20:PU138598 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Distribution-Based Global Sensitivity Analysis By Polynomial Chaos Expansion

Popis výsledku v původním jazyce

The paper is focused on study of distribution based global sensitivity indices derived directly from polynomial chaos expansion. The significant advantage is that, once the approximation in form of polynomial chaos expansion is available it is possible to obtain first statistical moments, Sobol indices and also distribution function with proposed moment-independent sensitivity indices without additional computational demands. The key idea is to use only specific part of approximation and compare obtained conditional probability cumulative distribution function to original distribution assuming all variables free to vary. The difference between distributions is measured by Cramer-von Misses distance herein. However, it is generally possible to employ any type of measure. The method is validated by analytical example with known solution. Proposed approach is highly efficient and thus it can be recommended for practical applications, when it is not possible to perform sensitivity analysis by standard Monte Carlo approach.

Název v anglickém jazyce

Distribution-Based Global Sensitivity Analysis By Polynomial Chaos Expansion

Popis výsledku anglicky

The paper is focused on study of distribution based global sensitivity indices derived directly from polynomial chaos expansion. The significant advantage is that, once the approximation in form of polynomial chaos expansion is available it is possible to obtain first statistical moments, Sobol indices and also distribution function with proposed moment-independent sensitivity indices without additional computational demands. The key idea is to use only specific part of approximation and compare obtained conditional probability cumulative distribution function to original distribution assuming all variables free to vary. The difference between distributions is measured by Cramer-von Misses distance herein. However, it is generally possible to employ any type of measure. The method is validated by analytical example with known solution. Proposed approach is highly efficient and thus it can be recommended for practical applications, when it is not possible to perform sensitivity analysis by standard Monte Carlo approach.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20102 - Construction engineering, Municipal and structural engineering

Projekt

<a href="/cs/project/GA20-01734S" target="_blank" >GA20-01734S: Pravděpodobnostně orientovaná globální citlivostní měření konstrukční spolehlivosti</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

ENGINEERING MECHANICS 2018 PROCEEDINGS, VOL 26

ISBN

978-80-214-5896-3

ISSN

—

e-ISSN

—

Počet stran výsledku

4

Strana od-do

380-383

Název nakladatele

Institute of Theretical and Applied Mechanics of the Czech Academy of Sciences

Místo vydání

Praha

Místo konání akce

Online

Datum konání akce

24. 11. 2020

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

000667956100087