Probabilistic Approach to Machine-Components Grouping in Celluar Manufacturing Systems

Abstract

In this paper existing group technology techniques are reviewed and an alternative method using probabilistic approach to machine-components grouping in cellular manufacturing systems is introduced where it is based on production flow analysis, which uses routing information. A common feature of this approach is that it sequentially rearranges row and columns of the machine part incidence matrix according to predefined index and block diagonal is generated. The steps of this method are to assign the 1's in each row and column a probability weight, which alternately rearranged in descending order until a block diagonal matrix is created. It does not need to decide in advance the number of required cells. It also overcomes the limitation of computational complexity, inherited in exiting group technology methods, especially for large scale and complex problems.