000004514 001__ 4514
000004514 005__ 20141118192702.0
000004514 0177_ $$2doi$$a10.3850/978-981-07-2219-7_P206

000004514 0247_ $$210.3850/978-981-07-2219-7_P206
$$adoi
000004514 04107 $$aeng
000004514 046__ $$k2012-05-23
000004514 100__ $$aNakatsu, Koichiro
000004514 24500 $$aReliability Analysis using Evolutionary Algorithm and Importance Sampling

000004514 24630 $$n5.$$pProceedings of the 5th Asian-Pacific Symposium on Structural Reliability and its Applications
000004514 260__ $$bResearch Publishing, No:83 Genting Lane, #08-01, Genting Building, 349568 SINGAPORE
000004514 506__ $$arestricted
000004514 520__ $$2eng$$aThe purpose of this research is the development of the method for reliability analysis to largescale structures. In large-scale structures, the reliability analysis is required to consider various failure samples in order to improve the calculation accuracy of failure probability. However, it is difficult due to the trade-off between the calculation accuracy and the calculation time. In this study, an attempt is made to develop a new simulation method for calculating low failure probabilities by using both evolutionary algorithm and importance sampling. Evolutionary algorithm can obtain various solutions in the search space, while importance sampling is a variance reduction technique, the behind idea of which is that certain values of the input random variables in a simulation have more impact on the parameter being estimated than others. At first, the proposed method adopts the evolutionary algorithm in order to obtain various failure samples included in different failure modes. Then, the failure probability is calculated by using importance sampling to failure samples obtained. Several numerical examples were presented to demonstrate that the evolutionary algorithm could obtain various failure samples in the proposed method. On the other hand, the calculation accuracy of failure probability was not improved by applying obtained failure samples to Importance Sampling. The safety of structure cannot be improved by only the calculation of failure probability which has the high accuracy. Therefore, it is necessary to develop the reliability analysis using various failure samples obtained by the proposed method.

000004514 540__ $$aText je chráněný podle autorského zákona č. 121/2000 Sb.
000004514 653__ $$aReliability analysis, Large-scale structure, Low failure probability, Importance sampling, Evolutionary algorithm, Markov chain monte carlo.

000004514 7112_ $$a5th Asian-Pacific Symposium on Structural Reliability and its Applications$$cSingapore (SG)$$d2012-05-23 / 2012-05-25$$gAPSSRA2012
000004514 720__ $$aNakatsu, Koichiro$$iFuruta, Hitoshi$$iNomura, Yasutoshi$$iTakahashi, Kyosuke$$iIshibashi, Ken$$iKondo, Takuya
000004514 8560_ $$ffischerc@itam.cas.cz
000004514 8564_ $$s292882$$uhttps://invenio.itam.cas.cz/record/4514/files/P206.pdf$$yOriginal version of the author's contribution as presented on CD, .
000004514 962__ $$r4180
000004514 980__ $$aPAPER