Abstract
Machine selection problem, which is one of the most popular and important issues, has been widely studied due to its extensive applications in real world, such as manufacturing and semiconductor industry. In this study, we focus on how to select different types of machines to each workstation can the makespan be minimized under a limited cost. Besides, as the number of decision variables increases, it would become increasingly difficult to find the optimal solution. In this study, we propose a simulation optimization method, which combines nested partitions and factor screening method to solve the large-scale machine selection problems. By adopting factor screening method, we can identify the ranking of the importance of different workstations; therefore, the process of optimization can follow this ranking to find the optimal solution. As a result, the required computations can be efficiently reduced. Based on experimental results, the proposed method is able to find the nearly optimal solution or optimal solution within a limited computational budget. Furthermore, the method proposed in this study significantly outperforms the two existing algorithms.