Long term extreme analysis of FPSO mooring systems based on Kriging metamodel

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

3 Citations (Scopus)

Abstract

Establishing statistical distributions of the response extremes is of particular importance for the design of FPSO mooring systems and the related riser design. Long term time domain simulation is the most accurate design approach to determine the extreme responses. It involves, however, coupled dynamic analysis of FPSO mooring system for a large number of sea states and consequently the task is often prohibitively time consuming. To solve this problem, an approach for the long term extreme analysis based on a metamodel in conjunction with the design of experiment methodology is proposed in the paper. In this approach, Latin Hypercube Sampling (LHS) based on the design of experiment method, is performed to select a sub-set of sea states from all sea states. Short term distributions for this sub-set of sea states are simulated and estimated. Kriging metamodel, which can map the relations between the sea states characteristics and the short term distribution parameters, is then applied. The accuracy of the metamodel is investigated. The long term response distribution of moored FPSO systems for all sea states can be predicted based on the metamodel. This approach for the long term extreme analysis of FPSO mooring systems avoids the response analysis over all sea states and can greatly improve the computational efficiency of the long term extreme analysis of FPSO mooring systems.

LanguageEnglish
Title of host publicationProceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering
Number of pages10
Volume1B
DOIs
Publication statusPublished - 2014
EventASME 2014 33rd International Conference on Ocean, Offshore and Arctic Engineering, OMAE 2014 - San Francisco, United Kingdom
Duration: 8 Jun 201413 Jun 2014

Conference

ConferenceASME 2014 33rd International Conference on Ocean, Offshore and Arctic Engineering, OMAE 2014
CountryUnited Kingdom
CitySan Francisco
Period8/06/1413/06/14

Fingerprint

Mooring
Design of experiments
Computational efficiency
Dynamic analysis
Floating production storage and offloading
Sampling

Keywords

  • computational efficiency
  • design
  • experiments
  • interpolation
  • mathematical models
  • mooring
  • ocean currents
  • time domain analysis
  • FPSO

Cite this

Wang, A., Huang, S., & Barltrop, N. (2014). Long term extreme analysis of FPSO mooring systems based on Kriging metamodel. In Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering (Vol. 1B). [V01BT01A052] https://doi.org/10.1115/OMAE2014-24609
Wang, Aijun ; Huang, Shan ; Barltrop, Nigel. / Long term extreme analysis of FPSO mooring systems based on Kriging metamodel. Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering. Vol. 1B 2014.
@inproceedings{384cd46775f2403e9ebd42beb8f48d0c,
title = "Long term extreme analysis of FPSO mooring systems based on Kriging metamodel",
abstract = "Establishing statistical distributions of the response extremes is of particular importance for the design of FPSO mooring systems and the related riser design. Long term time domain simulation is the most accurate design approach to determine the extreme responses. It involves, however, coupled dynamic analysis of FPSO mooring system for a large number of sea states and consequently the task is often prohibitively time consuming. To solve this problem, an approach for the long term extreme analysis based on a metamodel in conjunction with the design of experiment methodology is proposed in the paper. In this approach, Latin Hypercube Sampling (LHS) based on the design of experiment method, is performed to select a sub-set of sea states from all sea states. Short term distributions for this sub-set of sea states are simulated and estimated. Kriging metamodel, which can map the relations between the sea states characteristics and the short term distribution parameters, is then applied. The accuracy of the metamodel is investigated. The long term response distribution of moored FPSO systems for all sea states can be predicted based on the metamodel. This approach for the long term extreme analysis of FPSO mooring systems avoids the response analysis over all sea states and can greatly improve the computational efficiency of the long term extreme analysis of FPSO mooring systems.",
keywords = "computational efficiency, design, experiments, interpolation, mathematical models, mooring, ocean currents, time domain analysis, FPSO",
author = "Aijun Wang and Shan Huang and Nigel Barltrop",
year = "2014",
doi = "10.1115/OMAE2014-24609",
language = "English",
isbn = "9780791845387",
volume = "1B",
booktitle = "Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering",

}

Wang, A, Huang, S & Barltrop, N 2014, Long term extreme analysis of FPSO mooring systems based on Kriging metamodel. in Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering. vol. 1B, V01BT01A052, ASME 2014 33rd International Conference on Ocean, Offshore and Arctic Engineering, OMAE 2014, San Francisco, United Kingdom, 8/06/14. https://doi.org/10.1115/OMAE2014-24609

Long term extreme analysis of FPSO mooring systems based on Kriging metamodel. / Wang, Aijun; Huang, Shan; Barltrop, Nigel.

Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering. Vol. 1B 2014. V01BT01A052.

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

TY - GEN

T1 - Long term extreme analysis of FPSO mooring systems based on Kriging metamodel

AU - Wang, Aijun

AU - Huang, Shan

AU - Barltrop, Nigel

PY - 2014

Y1 - 2014

N2 - Establishing statistical distributions of the response extremes is of particular importance for the design of FPSO mooring systems and the related riser design. Long term time domain simulation is the most accurate design approach to determine the extreme responses. It involves, however, coupled dynamic analysis of FPSO mooring system for a large number of sea states and consequently the task is often prohibitively time consuming. To solve this problem, an approach for the long term extreme analysis based on a metamodel in conjunction with the design of experiment methodology is proposed in the paper. In this approach, Latin Hypercube Sampling (LHS) based on the design of experiment method, is performed to select a sub-set of sea states from all sea states. Short term distributions for this sub-set of sea states are simulated and estimated. Kriging metamodel, which can map the relations between the sea states characteristics and the short term distribution parameters, is then applied. The accuracy of the metamodel is investigated. The long term response distribution of moored FPSO systems for all sea states can be predicted based on the metamodel. This approach for the long term extreme analysis of FPSO mooring systems avoids the response analysis over all sea states and can greatly improve the computational efficiency of the long term extreme analysis of FPSO mooring systems.

AB - Establishing statistical distributions of the response extremes is of particular importance for the design of FPSO mooring systems and the related riser design. Long term time domain simulation is the most accurate design approach to determine the extreme responses. It involves, however, coupled dynamic analysis of FPSO mooring system for a large number of sea states and consequently the task is often prohibitively time consuming. To solve this problem, an approach for the long term extreme analysis based on a metamodel in conjunction with the design of experiment methodology is proposed in the paper. In this approach, Latin Hypercube Sampling (LHS) based on the design of experiment method, is performed to select a sub-set of sea states from all sea states. Short term distributions for this sub-set of sea states are simulated and estimated. Kriging metamodel, which can map the relations between the sea states characteristics and the short term distribution parameters, is then applied. The accuracy of the metamodel is investigated. The long term response distribution of moored FPSO systems for all sea states can be predicted based on the metamodel. This approach for the long term extreme analysis of FPSO mooring systems avoids the response analysis over all sea states and can greatly improve the computational efficiency of the long term extreme analysis of FPSO mooring systems.

KW - computational efficiency

KW - design

KW - experiments

KW - interpolation

KW - mathematical models

KW - mooring

KW - ocean currents

KW - time domain analysis

KW - FPSO

UR - http://www.scopus.com/inward/record.url?scp=84911430257&partnerID=8YFLogxK

UR - http://proceedings.asmedigitalcollection.asme.org/volume.aspx?conferenceid=3389&volumeid=16704

UR - http://www.asmeconferences.org/omae2014/

U2 - 10.1115/OMAE2014-24609

DO - 10.1115/OMAE2014-24609

M3 - Conference contribution book

SN - 9780791845387

VL - 1B

BT - Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering

ER -

Wang A, Huang S, Barltrop N. Long term extreme analysis of FPSO mooring systems based on Kriging metamodel. In Proceedings of the ASME 2014 33rd International Conference on Offshore Mechanics and Arctic Engineering. Vol. 1B. 2014. V01BT01A052 https://doi.org/10.1115/OMAE2014-24609