Machine-cell formation through neural network models

Abstract

Identification of machine-cells is one of the most important problems in the design of cellular manufacturing systems (CMSs). It involves decomposing a manufacturing system into machine-cells by grouping machines and parts. Several algorithms with varying degrees of success have been proposed and utilised to solve this problem. Among the modern tools, neural network models have the potential to solve the machine-cell formation problem. Choosing the competitive learning model, adaptive resonance theory (ART) model and self-organizing feature map (SOFM) model from neural network theory for this purpose, we demonstrate their suitability for solving the machine-cell formation problem. Applications on trial problems show the viability for solving the machine-cell formations problem and stand testimony to the practical utility of neural network models in designing CMSs.