Efficient Methodology for High Level Decision Making on a Manufacturing Facility

Abstract

The application of artificial intelligence methodologies for the development of decision support systems in the manufacturing field has led to promising results. Nevertheless, the combinatorial explosion in decision problems related to systems with complex behavior requires a large use of computational resources, which may imply that many developed methodologies are unpractical for real-time applications. In this paper, the choice of the best production strategy for a manufacturing facility is afforded by means of an improved methodology, based in modeling with the formalism of the disjunctive colored Petri nets and the application of a search process in the solution space by means of a metaheuristic. A genetic algorithm has been chosen as an adequate artificial intelligence technique for solving this high level decision making. A classic approach has also been applied to this manufacturing facility to compare its performance with the proposed methodology. The new approach outperforms the classic one in this case study.