The cell formation problem is a crucial component of a cell production design in a manufacturing system. Problems related to the cell formation problem are complex NP-hard problems. The goal of the work is to design the algorithm for the cell formation problem that is more efficient then the best-known algorithms for the same problem. The strategy of the new approach is to use the specificities of the input instances to narrow down the feasible set, and thus increase the efficiency of the optimization process. In the dynamic production environment, efficacy is one of the most significant characteristics of the applied expert system. The result is, extensible hybrid algorithm that can be used to solve complex, multi-criteria optimization cell formation problems. The new algorithm produces solutions that are as good as, or better than, the best results previously reported in literature on all commonly used test instances. The time efficiency of the proposed algorithm is at least an order of magnitude better than the efficiency of the most efficient reported algorithms. The obtained experimental results, modularity and generality of the new algorithm imply the significant impact on the expert systems for cell formation problem since the proposed strategy can improve the efficiency of existing algorithms for the grouping problems.

Introduction

Group Technology (GT) is defined as an approach to work optimization in which the organizational production units are relatively independent. In other words, GT is a manufacturing philosophy that exploits similarities in product design and product processes. Products needing similar operations and a common set of resources are grouped into families, the resources being regrouped into production subsystems, cells. The major advantages of cellular manufacturing reported in the literature include: reduction in setup time, reduction in labor and overtime costs, reduction in work-in-process inventories, reduction in material handling costs, and faster response to internal and external changes such as: machine failures, product mix, and demand changes (Wu, Chang, & Yeh, 2009). In the globalized and interconnected market, a production system must have a high degree of flexibility and agility to deal with product changes. The Dynamic Cellular Manufacturing System (DCMS) is one of the well-known production systems that meet this requirement. Problems related to DCMS are complex and time-consuming NP-hard problems in their nature. The memory and computational time requirements are extremely high, and increase exponentially, as the problem size increases. In the dynamic production environment, a multi-period planning horizon is considered where each period has a different product mix and demand necessities. Therefore, the formed cells in a period may not be optimal for the next period. To address this production environment, two approaches could be implemented. One is the concept of virtual cells, initially introduced by (McLean, Bloom, & Hopp, 1982), which allows exploiting the processing of part families into machines grouped virtually.