A METAHEURISTIC APPROACH FOR GENERALIZED CELL FORMATION WITH RATIO-LEVEL DATA

Abstract: Cellular layout is a well-known solution method to the layout problems in modern batch type production industries. In Cellular manufacturing system (CMS), machines and parts are grouped together based on manufacturing similarities. Generally, zero-one machine-part incidence matrix (MPIM) obtained from the route sheet information is used to form machine cells. The major demerit with such approaches is that real life production factors cannot be accounted for. In an attempt to address generalized cell formation problem, this study considers operational time of the parts processed on each machine. An efficient algorithm based on Genetic Algorithm (GA) is suggested to form machine cells with the combined objective of minimizing total cell load variation and exceptional elements when workload matrix is presented as input. A new grouping efficiency measure called Modified Grouping Efficiency (MGE) is proposed to evaluate the performance of GA. The proposed algorithm is tested with varied sizes of problems from open literature