Comparison of Adaptively Updated Surrogate Models for Reliability Analysis

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21110%2F14%3A00219435" target="_blank" >RIV/68407700:21110/14:00219435 - isvavai.cz</a>

Výsledek na webu

<a href="http://www.esco2014.femhub.com/docs/ESCO2014_Book_of_Abstracts.pdf" target="_blank" >http://www.esco2014.femhub.com/docs/ESCO2014_Book_of_Abstracts.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Comparison of Adaptively Updated Surrogate Models for Reliability Analysis

Popis výsledku v původním jazyce

Meta-models (or surrogate models, formerly response surfaces) are getting popular in engineering designs. They are used to simulate the behaviour of structures with less computational demands than the original model (e.g. finite element models). It is still necessary to evaluate this expensive original model few times in some specified points called Design of Experiments (DoE). The first DoE is usually just space-filling to cover up the whole design space and the meta-model built with an initial DoE istherefore not very accurate everywhere. Recently, a new updating procedure of a general surrogate model has been proposed. New points are added to the meta-model to improve accuracy at important regions based on a result of multi-objective optimization.There are two criteria: first, a new support point should be in the vicinity of the limit state function. Second, to bring the maximum new information, the point should be far from other points, which is performed by appropriate distance

Název v anglickém jazyce

Comparison of Adaptively Updated Surrogate Models for Reliability Analysis

Popis výsledku anglicky

Meta-models (or surrogate models, formerly response surfaces) are getting popular in engineering designs. They are used to simulate the behaviour of structures with less computational demands than the original model (e.g. finite element models). It is still necessary to evaluate this expensive original model few times in some specified points called Design of Experiments (DoE). The first DoE is usually just space-filling to cover up the whole design space and the meta-model built with an initial DoE istherefore not very accurate everywhere. Recently, a new updating procedure of a general surrogate model has been proposed. New points are added to the meta-model to improve accuracy at important regions based on a result of multi-objective optimization.There are two criteria: first, a new support point should be in the vicinity of the limit state function. Second, to bring the maximum new information, the point should be far from other points, which is performed by appropriate distance

Druh

O - Ostatní výsledky

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

<a href="/cs/project/GAP105%2F12%2F1146" target="_blank" >GAP105/12/1146: Metody paralelizace inženýrských úloh využívající cenově dostupné technologie</a><br>

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2014

Kód důvěrnosti údajů

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