ANN APPROACH TO THE - SOLUTION OF CELL FORMATION PROBLEMS IN CELLULAR MANUFACTURING

Abstract

Cell formation is an important issue in the design of Cellular Manufacturing system. Cell formation consists of identifying machine groups and part-families. Consideration of alternate process plans for the parts enable the user to get the better Cell formation. An Artificial Neural Network based clustering method ART-1 (Adaptive Resonance Theory-1) is used for the solution of Cell formation problems. ART-1 artificial neural network •is selected because it can take binary valued inputs and it can be trained without supervision. In present work the author has developed a computer software to solve the Cell formation problems with or without alternate .process plans. This software is written in C language to be operable in DOS environment. The software deals with the following issues: (1.) Consider alternate process plans. (2.) Reduces the presence of voids and exceptional elements. (3.) Provide grouping of parts and machines. (4.) Provide flexibility to the designer to form large loose cells or small tight cells by taking different values of similarity coefficient (vigilance parameter).