MODELING AND SIMULATION OF GENETIC SUPERVISORY FUZZY CONTROLLERS FOR MULTI-PART-TYPE PRODUCTION LINE
Seyed Mahdi Homayouni, N.H.F. Ismail · 2007
Genetic supervisory fuzzy (GSF) control architecture for multi-part-type production line is proposed. More than one part type can be processed by each machine in such production systems. GSF control architecture composed of two layer controller. In first layer heuristic distributed fuzzy (HDF) controllers control each machine separately, while GSF controllers used in second layer. GSF controllers tune the decisions made by HDF controllers, based on overall conditions of production system. Genetic algorithm (GA) is used to adapt the membership functions of supervisory fuzzy controllers, to improve the performance of GSF controllers. The overall objective is to control the production rate in a way that satisfies the demand for final products while keeping minimum work-in-process (WIP) and backlog within the production system. GA is used to minimize costs of WIP and backlog. The GSF control architecture is tested and compared with the heuristic supervisory fuzzy (HSF) controllers. The results show that in most of the cases GSF outperform the conventional supervisory controllers.