EXPERIMENTAL DESIGN ISSUES FOR SIMULTANEOUS FITTING OF FORWARD AND INVERSE METAMODELS

Russell R. Barton, Martin Meckesheimer, Timothy William Simpson

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

3 Scopus citations

Abstract

In engineering design, approximations to engineering simulation or analysis models can provide quick, approximate predictions of system or part performance as a function of system/part design parameters. There are often preliminary requirements for performance, which can be mapped to specific values of design parameters via an inverse model. Typically, only forward models exist explicitly, but an inverse approximation can be fitted. The quality of an approximation depends on the set of runs (the experiment design) used to fit the approximation. This paper examines experiment design construction strategies for simultaneously fitting forward and inverse approximation models, using the commonly applied measure of D-optimality.

Original languageEnglish (US)
Title of host publication26th Design Automation Conference
PublisherAmerican Society of Mechanical Engineers (ASME)
Pages773-780
Number of pages8
ISBN (Electronic)9780791835128
DOIs
StatePublished - 2000
EventASME 2000 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2000 - Baltimore, United States
Duration: Sep 10 2000Sep 13 2000

Publication series

NameProceedings of the ASME Design Engineering Technical Conference
Volume2

Conference

ConferenceASME 2000 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2000
Country/TerritoryUnited States
CityBaltimore
Period9/10/009/13/00

All Science Journal Classification (ASJC) codes

  • Mechanical Engineering
  • Computer Graphics and Computer-Aided Design
  • Computer Science Applications
  • Modeling and Simulation

Fingerprint

Dive into the research topics of 'EXPERIMENTAL DESIGN ISSUES FOR SIMULTANEOUS FITTING OF FORWARD AND INVERSE METAMODELS'. Together they form a unique fingerprint.

Cite this