Manufacturing cell formation by competitive learning

Cell formation (CF), one of the first problems faced in designing a cellular manufacturing system, involves grouping similar parts into families and corresponding machines into cells. The problem has attracted much attention from academia and industries, and an extensive amount of effort has been put into developing efficient procedures. In this paper, a neural network approach based upon a competitive learning paradigm is proposed to handle such a problem. Computational experience shows that the procedure is fairly efficient and can effectively obtain optimal clustering results.