Uncertainty management in multidisciplinary design of critical safety systems

Edoardo Patelli, Diego A. Alvarez, Matteo Broggi, Marco de Angelis

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Managing the uncertainty in multidisciplinary design of safety-critical systems requires not only the availability of a single approach or methodology to deal with uncertainty but a set of different strategies and scalable computational tools (that is, by making use of the computational power of a cluster and grid computing). The availability of multiple tools and approaches for dealing with uncertainties allows cross validation of the results and increases the confidence in the performed analysis. This paper presents a unified theory and an integrated and open general-purpose computational framework to deal with scarce data, and aleatory and epistemic uncertainties. It allows solving of the different tasks necessary to manage the uncertainty, such as uncertainty characterization, sensitivity analysis, uncertainty quantification, and robust design. The proposed computational framework is generally applicable to solve different problems in different fields and be numerically efficient and scalable, allowing for a significant reduction of the computational time required for uncertainty management and robust design. The applicability of the proposed approach is demonstrated by solving a multidisciplinary design of a critical system proposed by NASA Langley Research Center in the multidisciplinary uncertainty quantification challenge problem.
LanguageEnglish
Pages140-169
Number of pages30
JournalJournal of Aerospace Information Systems
Volume12
Issue number1
DOIs
Publication statusPublished - 31 Jan 2015

Fingerprint

Security systems
Availability
Cluster computing
Uncertainty
Grid computing
Sensitivity analysis
NASA

Keywords

  • uncertainty management
  • critical safety systems
  • multidisciplinary design
  • computational tools

Cite this

Patelli, Edoardo ; Alvarez, Diego A. ; Broggi, Matteo ; Angelis, Marco de. / Uncertainty management in multidisciplinary design of critical safety systems. In: Journal of Aerospace Information Systems. 2015 ; Vol. 12, No. 1. pp. 140-169.
@article{d28b221a3fd04cf1b160d99f802ca973,
title = "Uncertainty management in multidisciplinary design of critical safety systems",
abstract = "Managing the uncertainty in multidisciplinary design of safety-critical systems requires not only the availability of a single approach or methodology to deal with uncertainty but a set of different strategies and scalable computational tools (that is, by making use of the computational power of a cluster and grid computing). The availability of multiple tools and approaches for dealing with uncertainties allows cross validation of the results and increases the confidence in the performed analysis. This paper presents a unified theory and an integrated and open general-purpose computational framework to deal with scarce data, and aleatory and epistemic uncertainties. It allows solving of the different tasks necessary to manage the uncertainty, such as uncertainty characterization, sensitivity analysis, uncertainty quantification, and robust design. The proposed computational framework is generally applicable to solve different problems in different fields and be numerically efficient and scalable, allowing for a significant reduction of the computational time required for uncertainty management and robust design. The applicability of the proposed approach is demonstrated by solving a multidisciplinary design of a critical system proposed by NASA Langley Research Center in the multidisciplinary uncertainty quantification challenge problem.",
keywords = "uncertainty management, critical safety systems, multidisciplinary design, computational tools",
author = "Edoardo Patelli and Alvarez, {Diego A.} and Matteo Broggi and Angelis, {Marco de}",
year = "2015",
month = "1",
day = "31",
doi = "10.2514/1.I010273",
language = "English",
volume = "12",
pages = "140--169",
journal = "Journal of Aerospace Information Systems",
issn = "2327-3097",
number = "1",

}

Uncertainty management in multidisciplinary design of critical safety systems. / Patelli, Edoardo; Alvarez, Diego A.; Broggi, Matteo; Angelis, Marco de.

In: Journal of Aerospace Information Systems, Vol. 12, No. 1, 31.01.2015, p. 140-169.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Uncertainty management in multidisciplinary design of critical safety systems

AU - Patelli, Edoardo

AU - Alvarez, Diego A.

AU - Broggi, Matteo

AU - Angelis, Marco de

PY - 2015/1/31

Y1 - 2015/1/31

N2 - Managing the uncertainty in multidisciplinary design of safety-critical systems requires not only the availability of a single approach or methodology to deal with uncertainty but a set of different strategies and scalable computational tools (that is, by making use of the computational power of a cluster and grid computing). The availability of multiple tools and approaches for dealing with uncertainties allows cross validation of the results and increases the confidence in the performed analysis. This paper presents a unified theory and an integrated and open general-purpose computational framework to deal with scarce data, and aleatory and epistemic uncertainties. It allows solving of the different tasks necessary to manage the uncertainty, such as uncertainty characterization, sensitivity analysis, uncertainty quantification, and robust design. The proposed computational framework is generally applicable to solve different problems in different fields and be numerically efficient and scalable, allowing for a significant reduction of the computational time required for uncertainty management and robust design. The applicability of the proposed approach is demonstrated by solving a multidisciplinary design of a critical system proposed by NASA Langley Research Center in the multidisciplinary uncertainty quantification challenge problem.

AB - Managing the uncertainty in multidisciplinary design of safety-critical systems requires not only the availability of a single approach or methodology to deal with uncertainty but a set of different strategies and scalable computational tools (that is, by making use of the computational power of a cluster and grid computing). The availability of multiple tools and approaches for dealing with uncertainties allows cross validation of the results and increases the confidence in the performed analysis. This paper presents a unified theory and an integrated and open general-purpose computational framework to deal with scarce data, and aleatory and epistemic uncertainties. It allows solving of the different tasks necessary to manage the uncertainty, such as uncertainty characterization, sensitivity analysis, uncertainty quantification, and robust design. The proposed computational framework is generally applicable to solve different problems in different fields and be numerically efficient and scalable, allowing for a significant reduction of the computational time required for uncertainty management and robust design. The applicability of the proposed approach is demonstrated by solving a multidisciplinary design of a critical system proposed by NASA Langley Research Center in the multidisciplinary uncertainty quantification challenge problem.

KW - uncertainty management

KW - critical safety systems

KW - multidisciplinary design

KW - computational tools

U2 - 10.2514/1.I010273

DO - 10.2514/1.I010273

M3 - Article

VL - 12

SP - 140

EP - 169

JO - Journal of Aerospace Information Systems

T2 - Journal of Aerospace Information Systems

JF - Journal of Aerospace Information Systems

SN - 2327-3097

IS - 1

ER -