A sampling-based RBDO algorithm with local refinement and efficient gradient estimation

Sylvain Lacaze, Samy Missoum, Loïc Brevault, Mathieu Balesdent

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This article describes a two stage Reliability-Based Design Optimization (RBDO) algorithm. The first stage consists of solving an approximated RBDO problem using meta-models. In order to use gradient-based techniques, the sensitivity of failure probabilities are derived with respect to hyperparameters of random variables as well as, and this is a novelty, deterministic variables. The second stage focuses on the local refinement of the meta-models around the first stage solution using generalized "max-min" samples. The approach is demonstrated on three examples including a crashworthiness problem with 11 random variables and 10 probabilistic constraints.

Original languageEnglish (US)
Title of host publication12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015
PublisherUniversity of British Columbia
ISBN (Electronic)9780888652454
StatePublished - 2015
Event12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2012 - Vancouver, Canada
Duration: Jul 12 2015Jul 15 2015

Other

Other12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2012
CountryCanada
CityVancouver
Period7/12/157/15/15

Fingerprint

Gradient Estimation
Local Refinement
Efficient Estimation
Metamodel
Random variables
Optimization Algorithm
Random variable
Crashworthiness
Sampling
Probabilistic Constraints
Hyperparameters
Failure Probability
Generalized Solution
Min-max
Gradient
Optimization Problem
Design optimization

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Statistics and Probability

Cite this

Lacaze, S., Missoum, S., Brevault, L., & Balesdent, M. (2015). A sampling-based RBDO algorithm with local refinement and efficient gradient estimation. In 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015 University of British Columbia.

A sampling-based RBDO algorithm with local refinement and efficient gradient estimation. / Lacaze, Sylvain; Missoum, Samy; Brevault, Loïc; Balesdent, Mathieu.

12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015. University of British Columbia, 2015.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lacaze, S, Missoum, S, Brevault, L & Balesdent, M 2015, A sampling-based RBDO algorithm with local refinement and efficient gradient estimation. in 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015. University of British Columbia, 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2012, Vancouver, Canada, 7/12/15.
Lacaze S, Missoum S, Brevault L, Balesdent M. A sampling-based RBDO algorithm with local refinement and efficient gradient estimation. In 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015. University of British Columbia. 2015
Lacaze, Sylvain ; Missoum, Samy ; Brevault, Loïc ; Balesdent, Mathieu. / A sampling-based RBDO algorithm with local refinement and efficient gradient estimation. 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015. University of British Columbia, 2015.
@inproceedings{7964f820e84247ce8020fc8811bd4b80,
title = "A sampling-based RBDO algorithm with local refinement and efficient gradient estimation",
abstract = "This article describes a two stage Reliability-Based Design Optimization (RBDO) algorithm. The first stage consists of solving an approximated RBDO problem using meta-models. In order to use gradient-based techniques, the sensitivity of failure probabilities are derived with respect to hyperparameters of random variables as well as, and this is a novelty, deterministic variables. The second stage focuses on the local refinement of the meta-models around the first stage solution using generalized {"}max-min{"} samples. The approach is demonstrated on three examples including a crashworthiness problem with 11 random variables and 10 probabilistic constraints.",
author = "Sylvain Lacaze and Samy Missoum and Lo{\"i}c Brevault and Mathieu Balesdent",
year = "2015",
language = "English (US)",
booktitle = "12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015",
publisher = "University of British Columbia",
address = "Canada",

}

TY - GEN

T1 - A sampling-based RBDO algorithm with local refinement and efficient gradient estimation

AU - Lacaze, Sylvain

AU - Missoum, Samy

AU - Brevault, Loïc

AU - Balesdent, Mathieu

PY - 2015

Y1 - 2015

N2 - This article describes a two stage Reliability-Based Design Optimization (RBDO) algorithm. The first stage consists of solving an approximated RBDO problem using meta-models. In order to use gradient-based techniques, the sensitivity of failure probabilities are derived with respect to hyperparameters of random variables as well as, and this is a novelty, deterministic variables. The second stage focuses on the local refinement of the meta-models around the first stage solution using generalized "max-min" samples. The approach is demonstrated on three examples including a crashworthiness problem with 11 random variables and 10 probabilistic constraints.

AB - This article describes a two stage Reliability-Based Design Optimization (RBDO) algorithm. The first stage consists of solving an approximated RBDO problem using meta-models. In order to use gradient-based techniques, the sensitivity of failure probabilities are derived with respect to hyperparameters of random variables as well as, and this is a novelty, deterministic variables. The second stage focuses on the local refinement of the meta-models around the first stage solution using generalized "max-min" samples. The approach is demonstrated on three examples including a crashworthiness problem with 11 random variables and 10 probabilistic constraints.

UR - http://www.scopus.com/inward/record.url?scp=84978689402&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978689402&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84978689402

BT - 12th International Conference on Applications of Statistics and Probability in Civil Engineering, ICASP 2015

PB - University of British Columbia

ER -