Information fusion in industrial performance: a 2-additive Choquet-integral based approach

This study deals with the fusion of performance expressions in a manufacturing context, i.e. the mechanism which allows the computation of a global performance knowing the elementary ones. The characteristics of this kind of fusion is that the elementary performances to be combined are, on the one hand, often information of heterogeneous nature (cost, quality, delay...). On the other hand, they are associated to diversified and numerous objectives which interacts in different ways (negative or positive synergy, veto...). In this sense, the fuzzy subset theory provides tools to deal with: the heterogeneity of the entities involved, the commensurability of the elementary performances expressed in the interval [0,1], and the different behaviors of the fusion operation (compromise effect, optimistic or pessimistic effect...). Among all the fuzzy fusion operators, we consider here, the 2-additive Choquet fuzzy integral for modeling the different mutual interactions between the performance criteria and fusing their associated evaluations.