On the evolutionary-fuzzy control of WIP in manufacturing systems

The effectiveness of optimized fuzzy controllers in the production scheduling has been demonstrated in the past, through extensive use of evolutionary algorithms (EAs). The EA strategy tunes a set of distributed and supervisory control modules, whose objective is to control the production rate in a way that satisfies the demand for final products, while reducing WIP within the production system. The EA identifies optimal design solutions in a given search space. How robust and generic is the outcome of the evolutionary process? This paper faces this question by testing the evolutionary tuned fuzzy controllers under varying demand conditions assuming that the actual demand might be different to the one used for evolution. Extensive simulations of distributed and supervisory controllers show that the evolutionary-fuzzy strategy achieved a significant reduction of WIP in all production lines and networks tested.