An evolutionary approach of multicriteria manufacturing cell formation

Several approaches to designing cellular manufacturing systems exist, but they are generally based on the optimization of a single criterion, such as the minimization of the inter-cell movement or the total cost. Unfortunately, in practice, it is often necessary to take into account multiple criteria. We present an approach based on a niched Pareto evolutionary algorithm which does not rely on weights or distances to a target solution. The suggested method allows several solutions to be provided, so that a choice is offered to the decision maker, according to his/her particular constraints. A new recombination operator is also presented. This approach is illustrated on a cell formation problem using two criteria: the cell workload and the traffic between cells.