Uninorms in fuzzy systems modeling (Q5947551)

In fuzzy logic control and fuzzy expert systems, the problem of finding the value of the output associated with a particular value of the input variable, is referred to as the fuzzy modelling inference process. Here, it is shown that this process involves an aggregation step in which the contributions of the different rules of the fuzzy systems model are combined. In fuzzy control, the output of the model is built up, starting from the null set, by disjuncting weighted consequents of each of the rules. The weights are related to the relevancy of the rule to the particular input value being considered. In fuzzy expert systems, the author forms a conjunction of weighted consequents of each of the rules. Here again, the weights are related to the relevancy of the rule to the particular input value under consideration, and this relevancy itself is determined by the membership grade of the input value in antecedent fuzzy set of the rule.