A Simulation-Optimization strategy to deal simultaneously with tens of decision variables and multiple performance measures in manufacturing

This work addresses the multiple criteria simulation optimization problem. Such a problem entails using an optimization strategy to manipulate the parameters of a simulation model to arrive at the best possible configurations in the presence of several performance measures in conflict. Pareto Efficiency conditions are used in an iterative framework based on experimental design and pairwise comparison. In particular, this work improves upon and replaces the use of Data Envelopment Analysis to determine the efficient frontier and replaces the use of a single-pass algorithm previously proposed by our research group. The results show a rapid convergence to a more precise characterization of the Pareto-efficient solutions. In addition, the capability of the method to deal with fifty decision variables simultaneously is demonstrated through a case study in the fine-tuning of a manufacturing line.