EVALUATION OF MEASUREMENT UNCERTAINTY IN TESTS REQUIRING NON-LINEAR REGRESSION

J. A. Sousa, I. M. Smith, A. B. Forbes, A. S. Ribeiro, D. F. Sousa

Abstract

This work is concerned with the evaluation of measurement uncertainty arising from the use of non-linear regression in testing. While modern standards require a measured value to be accompanied by a statement of its quality in the form of an associated uncertainty, for many applications existing work procedures make no reference to uncertainty. Such procedures should therefore be updated to take account of the requirement to provide uncertainty information. This paper considers an application that involves the use of non-linear regression and for which uncertainty evaluation does not constitute part of the current work procedure. An updated procedure is proposed and example results are presented.

Keywords

uncertainty, non-linear regression, modelling, least squares adjustment

Download

IMEKO-WC-2012-TC7-O5.pdf

DOI

-

IMEKO TC

TC7 - Measurement Science

Back to the proceedings