A hybrid genetic algorithm to optimize the printed circuit board assembly process

Abstract

Based on the analysis of assembly process for the chip-shooter machine, a hybrid genetic algorithm (GA) for printed circuit board (PCB) assembly optimization was presented to minimize the assembly time. During the assembly process, the speed of PCB table movement and turret rotation is inconstant; it is affected by placement sequence and component size. According to the size of component and component placement sequence, all the components are classified into key components and free components. Polychromatic set (PS) contour matrix is used to describe the classification and the relationship of key components and free components. Through appropriate coding method and genetic operation, GA is used to simultaneously optimize two complex combinatorial optimization problems which include component placement sequence and feeder arrangement. The contour matrix combines with generation initialization, calculation of fitness value and genetic operation. The search space of GA is obviously reduced, and the solution efficiency is improved. Compare with other algorithm, hybrid GA can optimize the assembly time accurately and effectively.