Appling similarity coefficient based on euclidean distance matrix in cellular manufacturing systems.

Numerous clustering algorithms are used in cellular manufacturing systems, which are extremely important to the problem of machine part group formation. Due to its adaptability and computing effectiveness, the similarity coefficient method is the most popular approach to solving this problem. Many ways of calculating similarity coefficients have been developed. Still, only a few consider the factor of operational sequencing with repeated operations, which can lead to more exceptional material intercell transfers. This paper proposed a similarity coefficient based on the Euclidian distance matrix SC/EDM. Computational experiments were conducted using ten benchmark problem sets selected from published works. The machine utilization, cell efficiency, grouping efficacy, and the number of exceptional elements were measured to improve the performance. Compared to published benchmark results, the SC/EDM presented in this paper achieves better or equal results. Machine utilization and Grouping efficacy improved by seven out of ten, Cell efficiency improved by seven out of ten, and finally, no exceptional elements improved by five out of ten problems taken. [ABSTRACT FROM AUTHOR]