Uncertainty Analysis Principles and Methods.

摘要

This document has been prepared to define and catalog standard mathematical concepts and methods used to estimate measurement uncertainty in test instrumentation systems. The analysis methods outlined in this document provide a comprehensive approach to estimating measurement uncertainty. Basic guidelines are presented for estimating the uncertainty for the following measurement processes: (1) Direct Measurements The value of the quantity or subject parameter is obtained directly by measurement and is not determined indirectly by computing its value from the measurement of other variables or quantities. (2) Multivariate Measurements The value of the quantity or subject parameter is based on measurements of more than one attribute or quantity. (3) Measurement Systems The value of the quantity or subject parameter is measured with a system comprised of component modules arranged in series. The structured, step-by-step analysis procedures described address the important aspects of identifying measurement process errors and estimating their uncertainty. Advanced topics cover uncertainty growth over time and the refinement of uncertainty estimates using Bayesian methods.