Premium
An investigation of noise factor effects in parameter design
Author(s) -
Parkinson D. B.
Publication year - 2002
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/qre.446
Subject(s) - constraint (computer aided design) , constant (computer programming) , design of experiments , set (abstract data type) , noise (video) , mathematics , estimation theory , control theory (sociology) , computer science , statistics , geometry , control (management) , artificial intelligence , image (mathematics) , programming language
Parameter design aims to determine nominal values of a set of design parameters so as to minimize variability in one or more key performance measures, in the presence of uncertainties in the design parameters, whilst maintaining the required nominal (design point) performance and the overall design concept. A theoretical analysis of some aspects of parameter design and of some related approximate methods is carried out and the results studied with the aim of determining the effect of changes in the magnitudes of the uncertainties in the design parameters on the resulting values of nominal parameter values in the optimum design. It is shown that there is reason to expect that, allowing for the effects of constraints, a given parameter design would be robust against overall changes, by a constant factor, in the parameter uncertainties but not in the presence of changes in individual parameter uncertainties, or by numerous such changes when the change factor varies between individual parameters. Some numerical results are obtained for problems in beam design and are shown to support the above assertions. The results imply that nominal values in a previously determined robust design may continue to be valid when parameter tolerances change, provided such changes are by the same factor for all free design parameters, i.e. those not fixed on a constraint boundary. Copyright © 2002 John Wiley & Sons, Ltd.