If several rules have the same variable in conclusion, they have to be fired in parallel and all their conclusion fuzzy sets have to be aggregated in one fuzzy set to be used as entry for others rules. Conclusion fuzzy sets are aggregated with the min operator to take into account uncertainty in the final fuzzy set. If n rules, fired in parallel, produce n fuzzy sets , the final fuzzy set is obtained by:
Nevertheless, if there is no consistency in the rules base the final fuzzy set has to be normalized [3] by: