Solving scheduling problems in PCB assembly and its optimization using ACO

Abstract

The focus of this paper is to schedule printed circuit board (PCB) assembly process while minimizing the mean flow time as well as work-in-process inventories. Here, problem for scheduling different types of PCBs on a single sequential pick-and-place automatic machine is considered in which the total number of different components required to process each type of PCBs exceeds the capacity of the feeder rack. The above objective is achieved through minimizing the number of feeder rack changes or component switches and sequential placement of components. A component switch refers to removal of one type of component from the feeder rack and a different type of component is placed on it, and may occur when changing to next type of PCB. In order to reduce component switches, group technology is applied following the counts of new components needed to add for successive group formation. Mathematical models are developed for PCB grouping and PCB group sequencing problem and integrated as multi-functional model to determine the optimal sequence of component placements. Ant colony optimization (ACO) technique is used to solve the proposed model and the results are compared with the different component grouping methods available in the literature.