A comparison between different methods for processing the random part of random-fuzzy variables representing measurement results

In the recent years, fuzzy variables and random-fuzzy variables have been proposed to represent the measurement results with their associated uncertainty. However, up to now, the different authors do not yet agree in the mathematical way fuzzy variables should be composed together, so that different approaches have been proposed. This paper compares these approaches, in order to find their advantages and disadvantages and shows a new proposal, that is supposed to overcome, hopefully, the disadvantages of the original ones.