Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022 |
| Herausgeber/-innen | Michael Beer, Enrico Zio, Kok-Kwang Phoon, Bilal M. Ayyub |
| Seiten | 508-514 |
| Seitenumfang | 7 |
| Publikationsstatus | Veröffentlicht - 2022 |
| Veranstaltung | 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022 - Hannover, Deutschland Dauer: 4 Sept. 2022 → 7 Sept. 2022 |
Publikationsreihe
| Name | Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022 |
|---|
Abstract
Both random and interval variables can coexist in a single reliability problem. Such cases could pose a serious challenge for existing reliability analysis methods. In this paper, we present a parallel active learning Kriging method for hybrid reliability analysis under both random and interval variables. The key contribution of the proposed method is developing a parallel active learning strategy that can identify a batch of points at eacn iteration, and hence parallel computing. This is achieved by proposing a new learning function, called pseudo weighted expected risk function (PWERF), which is based on the use of the expected risk function, an influence function and the joint probability function of basis random variables. Once a predefined stopping criterion is satisfied, the lower and upper-bounds of the failure probability can be estimated from the Kriging model as a surrogate for the true performance function. Two numerical examples are employed to demonstrate the performance of the proposed method in comparison with an existing method.
ASJC Scopus Sachgebiete
- Entscheidungswissenschaften (insg.)
- Managementlehre und Operations Resarch
- Ingenieurwesen (insg.)
- Sicherheit, Risiko, Zuverlässigkeit und Qualität
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022. Hrsg. / Michael Beer; Enrico Zio; Kok-Kwang Phoon; Bilal M. Ayyub. 2022. S. 508-514 (Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Random-interval hybrid reliability analysis by a parallel active learning Kriging method with a pseudo weighted expected risk function
AU - Liu, J.
AU - Dang, C.
AU - Beer, M.
PY - 2022
Y1 - 2022
N2 - Both random and interval variables can coexist in a single reliability problem. Such cases could pose a serious challenge for existing reliability analysis methods. In this paper, we present a parallel active learning Kriging method for hybrid reliability analysis under both random and interval variables. The key contribution of the proposed method is developing a parallel active learning strategy that can identify a batch of points at eacn iteration, and hence parallel computing. This is achieved by proposing a new learning function, called pseudo weighted expected risk function (PWERF), which is based on the use of the expected risk function, an influence function and the joint probability function of basis random variables. Once a predefined stopping criterion is satisfied, the lower and upper-bounds of the failure probability can be estimated from the Kriging model as a surrogate for the true performance function. Two numerical examples are employed to demonstrate the performance of the proposed method in comparison with an existing method.
AB - Both random and interval variables can coexist in a single reliability problem. Such cases could pose a serious challenge for existing reliability analysis methods. In this paper, we present a parallel active learning Kriging method for hybrid reliability analysis under both random and interval variables. The key contribution of the proposed method is developing a parallel active learning strategy that can identify a batch of points at eacn iteration, and hence parallel computing. This is achieved by proposing a new learning function, called pseudo weighted expected risk function (PWERF), which is based on the use of the expected risk function, an influence function and the joint probability function of basis random variables. Once a predefined stopping criterion is satisfied, the lower and upper-bounds of the failure probability can be estimated from the Kriging model as a surrogate for the true performance function. Two numerical examples are employed to demonstrate the performance of the proposed method in comparison with an existing method.
KW - failure probability bounds
KW - hybrid reliability analysis
KW - Kriging model
KW - parallel computing
UR - http://www.scopus.com/inward/record.url?scp=85202068676&partnerID=8YFLogxK
U2 - 10.3850/978-981-18-5184-1_MS-15-221-cd
DO - 10.3850/978-981-18-5184-1_MS-15-221-cd
M3 - Conference contribution
AN - SCOPUS:85202068676
SN - 9789811851841
T3 - Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022
SP - 508
EP - 514
BT - Proceedings of the 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022
A2 - Beer, Michael
A2 - Zio, Enrico
A2 - Phoon, Kok-Kwang
A2 - Ayyub, Bilal M.
T2 - 8th International Symposium on Reliability Engineering and Risk Management, ISRERM 2022
Y2 - 4 September 2022 through 7 September 2022
ER -