A Bayesian approach to the quantification of extremal responses in simulated dynamic structures

Research output: Contribution to journalArticle

Abstract

Prediction of the extremal responses of dynamic structures is a vital step in the risk management of offshore assets. Often when modelling structural response the outputs are dependent on covariates defined on a continuous input domain. We demonstrate a methodology to allow for continuous covariates in extremal modelling by building latent variable models, whereby output dependencies are incorporated by smooth processes in the latent parameters. This allows information from close-by input regions to be shared when forming inference at unseen inputs. We illustrate the methodology using a computational simulation of Floating Production Storage and Offloading (FPSO)vessel motions, modelled as functions of the peak wave period. We provide methodologies and diagnostics for the modelling of the time-domain maxima, quantiles, and threshold exceedance data. There are three contributions made by this research: a methodology to predict the extremal outputs from a time-domain simulator, with incorporation of continuous covariate knowledge; significant speed increase when using the developed methodology as a computational proxy to the simulator; and a framework for the probabilistic quantification of the output uncertainty of the extremal data.

Original languageEnglish
Pages (from-to)594-607
Number of pages14
JournalOcean Engineering
Volume182
DOIs
Publication statusPublished - 15 Jun 2019

Fingerprint

Simulators
Risk management
Floating production storage and offloading
Uncertainty

Cite this

@article{66721ad2778a47199a511325007f5da6,
title = "A Bayesian approach to the quantification of extremal responses in simulated dynamic structures",
abstract = "Prediction of the extremal responses of dynamic structures is a vital step in the risk management of offshore assets. Often when modelling structural response the outputs are dependent on covariates defined on a continuous input domain. We demonstrate a methodology to allow for continuous covariates in extremal modelling by building latent variable models, whereby output dependencies are incorporated by smooth processes in the latent parameters. This allows information from close-by input regions to be shared when forming inference at unseen inputs. We illustrate the methodology using a computational simulation of Floating Production Storage and Offloading (FPSO)vessel motions, modelled as functions of the peak wave period. We provide methodologies and diagnostics for the modelling of the time-domain maxima, quantiles, and threshold exceedance data. There are three contributions made by this research: a methodology to predict the extremal outputs from a time-domain simulator, with incorporation of continuous covariate knowledge; significant speed increase when using the developed methodology as a computational proxy to the simulator; and a framework for the probabilistic quantification of the output uncertainty of the extremal data.",
keywords = "Bayesian statistics, Dynamic structures, Extremal data, FPSO vessel motions, Latent variable model, Uncertainty quantification",
author = "Astfalck, {L. C.} and Cripps, {E. J.} and Hodkiewicz, {M. R.} and Milne, {I. A.}",
year = "2019",
month = "6",
day = "15",
doi = "10.1016/j.oceaneng.2019.04.035",
language = "English",
volume = "182",
pages = "594--607",
journal = "Ocean Engineering",
issn = "0029-8018",
publisher = "Pergamon",

}

TY - JOUR

T1 - A Bayesian approach to the quantification of extremal responses in simulated dynamic structures

AU - Astfalck, L. C.

AU - Cripps, E. J.

AU - Hodkiewicz, M. R.

AU - Milne, I. A.

PY - 2019/6/15

Y1 - 2019/6/15

N2 - Prediction of the extremal responses of dynamic structures is a vital step in the risk management of offshore assets. Often when modelling structural response the outputs are dependent on covariates defined on a continuous input domain. We demonstrate a methodology to allow for continuous covariates in extremal modelling by building latent variable models, whereby output dependencies are incorporated by smooth processes in the latent parameters. This allows information from close-by input regions to be shared when forming inference at unseen inputs. We illustrate the methodology using a computational simulation of Floating Production Storage and Offloading (FPSO)vessel motions, modelled as functions of the peak wave period. We provide methodologies and diagnostics for the modelling of the time-domain maxima, quantiles, and threshold exceedance data. There are three contributions made by this research: a methodology to predict the extremal outputs from a time-domain simulator, with incorporation of continuous covariate knowledge; significant speed increase when using the developed methodology as a computational proxy to the simulator; and a framework for the probabilistic quantification of the output uncertainty of the extremal data.

AB - Prediction of the extremal responses of dynamic structures is a vital step in the risk management of offshore assets. Often when modelling structural response the outputs are dependent on covariates defined on a continuous input domain. We demonstrate a methodology to allow for continuous covariates in extremal modelling by building latent variable models, whereby output dependencies are incorporated by smooth processes in the latent parameters. This allows information from close-by input regions to be shared when forming inference at unseen inputs. We illustrate the methodology using a computational simulation of Floating Production Storage and Offloading (FPSO)vessel motions, modelled as functions of the peak wave period. We provide methodologies and diagnostics for the modelling of the time-domain maxima, quantiles, and threshold exceedance data. There are three contributions made by this research: a methodology to predict the extremal outputs from a time-domain simulator, with incorporation of continuous covariate knowledge; significant speed increase when using the developed methodology as a computational proxy to the simulator; and a framework for the probabilistic quantification of the output uncertainty of the extremal data.

KW - Bayesian statistics

KW - Dynamic structures

KW - Extremal data

KW - FPSO vessel motions

KW - Latent variable model

KW - Uncertainty quantification

UR - http://www.scopus.com/inward/record.url?scp=85065814573&partnerID=8YFLogxK

U2 - 10.1016/j.oceaneng.2019.04.035

DO - 10.1016/j.oceaneng.2019.04.035

M3 - Article

VL - 182

SP - 594

EP - 607

JO - Ocean Engineering

JF - Ocean Engineering

SN - 0029-8018

ER -