Clustered parameters of calibrated models when considering both fidelity and robustness

Sez Atamturktur, Garrison Stevens, Yuting Cheng

Research output: Contribution to journalConference article

Abstract

In computer modeling, errors and uncertainties inevitably arise due to the mathematical idealization of physical processes stemming from insufficient knowledge regarding accurate model forms as well as the precise values of input parameters. While these errors and uncertainties are quantifiable, compensations between them can lead to multiple model forms and input parameter sets exhibiting a similar level of agreement with available experimental observations. Such nonuniqueness makes the selection of a single, best computer model (i.e. model form and values for its associate parameters) unjustifiable. Therefore, it becomes necessary to evaluate model performance based not only on the fidelity of the predictions to available experiments but also on a model’s ability to sustain such fidelity given the incompleteness of knowledge regarding the model itself, such an ability will herein be referred to as robustness. In this paper, the authors present a multiobjective approach to model calibration that accounts for not only the model’s fidelity to experiments but also its robustness to incomplete knowledge. With two conflicting objectives, the multi-objective model calibration results in a family of nondominated solutions exhibiting varying levels of fidelity and robustness effectively forming a Pareto front. The Pareto front solutions can be grouped depending on their nature of compromise between the two objectives, which can in turn help determine clusters in the parameter domain. The knowledge of these clusters can shed light on the nature of compensations as well as aid in the inference of uncertain input parameters. To demonstrate the feasibility and application of this new approach, we consider the computer model of a structural steel frame with uncertain connection stiffness parameters under static loading conditions.

Original languageEnglish (US)
Article numberA30
Pages (from-to)215-224
Number of pages10
JournalConference Proceedings of the Society for Experimental Mechanics Series
Volume3
DOIs
StatePublished - Jan 1 2015
Event2014 Annual Conference on Experimental and Applied Mechanics, SEM 2014 - Greenville, SC, United States
Duration: Jun 2 2014Jun 5 2014

Fingerprint

Calibration
Experiments
Stiffness
Steel
Uncertainty
Compensation and Redress

All Science Journal Classification (ASJC) codes

  • Engineering(all)
  • Computational Mechanics
  • Mechanical Engineering

Cite this

@article{406adda18b7241fd9691ff6c7a30130e,
title = "Clustered parameters of calibrated models when considering both fidelity and robustness",
abstract = "In computer modeling, errors and uncertainties inevitably arise due to the mathematical idealization of physical processes stemming from insufficient knowledge regarding accurate model forms as well as the precise values of input parameters. While these errors and uncertainties are quantifiable, compensations between them can lead to multiple model forms and input parameter sets exhibiting a similar level of agreement with available experimental observations. Such nonuniqueness makes the selection of a single, best computer model (i.e. model form and values for its associate parameters) unjustifiable. Therefore, it becomes necessary to evaluate model performance based not only on the fidelity of the predictions to available experiments but also on a model’s ability to sustain such fidelity given the incompleteness of knowledge regarding the model itself, such an ability will herein be referred to as robustness. In this paper, the authors present a multiobjective approach to model calibration that accounts for not only the model’s fidelity to experiments but also its robustness to incomplete knowledge. With two conflicting objectives, the multi-objective model calibration results in a family of nondominated solutions exhibiting varying levels of fidelity and robustness effectively forming a Pareto front. The Pareto front solutions can be grouped depending on their nature of compromise between the two objectives, which can in turn help determine clusters in the parameter domain. The knowledge of these clusters can shed light on the nature of compensations as well as aid in the inference of uncertain input parameters. To demonstrate the feasibility and application of this new approach, we consider the computer model of a structural steel frame with uncertain connection stiffness parameters under static loading conditions.",
author = "Sez Atamturktur and Garrison Stevens and Yuting Cheng",
year = "2015",
month = "1",
day = "1",
doi = "10.1007/978-3-319-15224-0_23",
language = "English (US)",
volume = "3",
pages = "215--224",
journal = "Conference Proceedings of the Society for Experimental Mechanics Series",
issn = "2191-5644",

}

Clustered parameters of calibrated models when considering both fidelity and robustness. / Atamturktur, Sez; Stevens, Garrison; Cheng, Yuting.

In: Conference Proceedings of the Society for Experimental Mechanics Series, Vol. 3, A30, 01.01.2015, p. 215-224.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Clustered parameters of calibrated models when considering both fidelity and robustness

AU - Atamturktur, Sez

AU - Stevens, Garrison

AU - Cheng, Yuting

PY - 2015/1/1

Y1 - 2015/1/1

N2 - In computer modeling, errors and uncertainties inevitably arise due to the mathematical idealization of physical processes stemming from insufficient knowledge regarding accurate model forms as well as the precise values of input parameters. While these errors and uncertainties are quantifiable, compensations between them can lead to multiple model forms and input parameter sets exhibiting a similar level of agreement with available experimental observations. Such nonuniqueness makes the selection of a single, best computer model (i.e. model form and values for its associate parameters) unjustifiable. Therefore, it becomes necessary to evaluate model performance based not only on the fidelity of the predictions to available experiments but also on a model’s ability to sustain such fidelity given the incompleteness of knowledge regarding the model itself, such an ability will herein be referred to as robustness. In this paper, the authors present a multiobjective approach to model calibration that accounts for not only the model’s fidelity to experiments but also its robustness to incomplete knowledge. With two conflicting objectives, the multi-objective model calibration results in a family of nondominated solutions exhibiting varying levels of fidelity and robustness effectively forming a Pareto front. The Pareto front solutions can be grouped depending on their nature of compromise between the two objectives, which can in turn help determine clusters in the parameter domain. The knowledge of these clusters can shed light on the nature of compensations as well as aid in the inference of uncertain input parameters. To demonstrate the feasibility and application of this new approach, we consider the computer model of a structural steel frame with uncertain connection stiffness parameters under static loading conditions.

AB - In computer modeling, errors and uncertainties inevitably arise due to the mathematical idealization of physical processes stemming from insufficient knowledge regarding accurate model forms as well as the precise values of input parameters. While these errors and uncertainties are quantifiable, compensations between them can lead to multiple model forms and input parameter sets exhibiting a similar level of agreement with available experimental observations. Such nonuniqueness makes the selection of a single, best computer model (i.e. model form and values for its associate parameters) unjustifiable. Therefore, it becomes necessary to evaluate model performance based not only on the fidelity of the predictions to available experiments but also on a model’s ability to sustain such fidelity given the incompleteness of knowledge regarding the model itself, such an ability will herein be referred to as robustness. In this paper, the authors present a multiobjective approach to model calibration that accounts for not only the model’s fidelity to experiments but also its robustness to incomplete knowledge. With two conflicting objectives, the multi-objective model calibration results in a family of nondominated solutions exhibiting varying levels of fidelity and robustness effectively forming a Pareto front. The Pareto front solutions can be grouped depending on their nature of compromise between the two objectives, which can in turn help determine clusters in the parameter domain. The knowledge of these clusters can shed light on the nature of compensations as well as aid in the inference of uncertain input parameters. To demonstrate the feasibility and application of this new approach, we consider the computer model of a structural steel frame with uncertain connection stiffness parameters under static loading conditions.

UR - http://www.scopus.com/inward/record.url?scp=84945568392&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84945568392&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-15224-0_23

DO - 10.1007/978-3-319-15224-0_23

M3 - Conference article

AN - SCOPUS:84945568392

VL - 3

SP - 215

EP - 224

JO - Conference Proceedings of the Society for Experimental Mechanics Series

JF - Conference Proceedings of the Society for Experimental Mechanics Series

SN - 2191-5644

M1 - A30

ER -