Sizing manufacturing cell machines based on the simulation and an expert system

Abstract

This article developed a new method of the manufacturing cell sizing according to an approach based on the simulation and expert system. This method takes into consideration the stochastic aspect which governs the production system, the production scales of the product family to be treated, the products arrival law, the adopted scheduling and the used optimization criteria. This study starts with the modeling of a manufacturing cell and the simulation of its functioning using the software "ARENA", which provides performance ratings. In a second stage, the expert system exploits these performance measures by a confrontation of the obtained results to the pre-defined objectives. This step permits to lead to a possible decision in order to size each of the machines to obtain a better manufacturing cell functioning. This process is repeated iteratively until the obtention of a cell having the performances which conform to the objectives already defined. The method represents a very useful and economical approach to predict the system size. Finally, applications of the developed method for the sizing of manufacturing cell machines permitted to get satisfactory results.