An entropy based group setup strategy for PCB assembly

Abstract

Group setup strategy exploits the PCB similarity in forming the families of boards to minimize makespan that is composed of two attributes, the setup time and the placement time. The component similarity of boards in families reduces the setup time between families meanwhile, the geometric similarity reduces the placement time of boards within families. Current group setup strategy considers the component similarity and the geometric similarity by giving equal weights or by considering each similarity sequentially. In this paper, we propose an improved group setup strategy which combines component similarity and geometric similarity simultaneously. The entropy method is used to determine the weight of each similarity by capturing the importance of each similarity in different production environments. Test results show that the entropy based group setup strategy outperforms existing group setup strategies.