A Bayesian Approach to Information Fusion for Evaluating the Measurement Uncertainty

The Bayesian approach to uncertainty evaluation is a classical example for information fusion. It is based on both, the knowledge about the measuring process and the input quantities. Appropriate probability density functions for the input quantities may be obtained by utilizing the principle of maximum information entropy and the Bayes theorem. The knowledge about the measurement process is represented by the so-called model equation which forms the basis for the fusion of all involved input quantities. Compared to the ISO-GUM procedure, the Bayesian approach to uncertainty evaluation does not have any restriction related to nonlinearity and determination of confidence intervals