The Mathematical Theory of Evidence and Measurement Uncertainty - Comparison of Measurement Results Expressed in Terms of Random-Fuzzy Variables

In the previous papers, it was shown how possibility distributions can be effectively employed to represent and propagate uncertainty in measurements. In particular, it was shown how the Random-Fuzzy variables (RFVs) can be effectively employed to represent a measurement result. The effects of the systematic and random contributions to uncertainty can be well identified in the RFV, and all confidence intervals at all confidence levels are provided, so that complete information about the measurement result is given. Moreover, this distinction also allows one to model the propagation of the systematic and the random contributions in two different ways, according to their different nature and different behavior when they combine.