Content area
Full text
Received Dec 8, 2017; Revised Apr 19, 2018; Accepted May 7, 2018
This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1. Introduction
Measurement uncertainty is an important parameter in measurement results. To give scientific and proper evaluation of measurement uncertainty is an important factor to guarantee the development of modern measuring science. It is over two decades since “Guide to the Expression of Uncertainty in Measurement (GUM)” was first released in 1993. Initially, it mainly served in the field of physical science. Now, it is widely used in chemistry, biology, medicine, forensic medicine, astronomy, and other fields. Measurement uncertainty plays an increasingly prominent role in modern science [1–3].
Some of the existing studies on measurement uncertainty evaluation are based on historical experience, experts’ opinion, and prior data and fail to take the real time measurement data into account [4, 5]. Some of the studies are based on the information of measured samples and fail to take the historical information of the measurement system into account [6, 7]. These methods can not reflect the most up-to-date status of the measurement system and thus have negative impact on the reliability and soundness of uncertainty evaluation results.
With the advances in computer technology, new uncertainty evaluation methods have been developed and applied. Using Bayesian method to evaluate measurement uncertainty is an important direction for modern uncertainty evaluation methods to progress. Based on Bayesian information fusion, the uncertainty evaluation method can fully integrate the prior and the current sample information. The prior distribution is determined by the historical information, and the posterior distribution is deduced by integrating prior distribution and the current sample data with the Bayesian model, so as to achieve both the evaluation and updating of uncertainty [8–15].
Prior distribution is an important part of Bayesian statistical model. Properly determining the prior distribution according to historical information is a key point for evaluating measurement uncertainty with Bayesian methods. Most researches on Bayesian prior distribution merely give a brief introduction to some prior distribution methods but fail to propose in-depth analysis of the advantages, limitations, and application scope of the methods [16]. In addition, few studies...