Optimization of the multi-objective dynamic cell formation problem using a scatter search approach

This paper addresses the dynamic cell formation (DCF) problem with multiple conflicting objectives. Recent researches have mainly focus on single-objective cell formation procedures that deals with the identification of part families and associated machine groups for constant demands. However, varying market demands and fluctuations of the business environment have caused cellular manufacturing systems to operate under dynamic conditions. Thus, the optimal configuration of manufacturing cells in each period is different and the reconfiguration of cells is required. This paper proposes a nonlinear multi-objective mathematical model of the DCF problem by giving weighing to three conflicting objectives including the machine relocation cost in the process of reconfiguring cells, the utilization rate of machine capacity, and the total number of intercell moves over the entire planning horizon. To solve the nonlinear multi-objective model, a scatter search approach is developed, which redesigns the common components of scatter search and incorporates diversification generator, global criterion method, local search method, and other improvement mechanisms to provide a wide exploration of the search space through intensification and diversification. The proposed approach is compared with the commercial solver CPLEX on 10 test problems, some of which are large dimensions. Computational results have demonstrated the effectiveness of the scatter search approach.