A genetic algorithm for sequential part assignment for PCB assembly

We present a genetic algorithm for printed circuit board (PCB) assembly, which simultaneously solves the feeder assignment and component sequencing problems. The algorithm uses a unique gene selection procedure that increases the convergence rate without degrading the solution quality. We have compared the performance of our algorithm with existing approaches and have demonstrated improved performance. We have also implemented the algorithm on a Quad IIIc insertion machine for surface mount components. The resulting suboptimal assembly times estimated by the algorithm have been found to be close to the actual optimal values. (C) 2001 Elsevier Science Ltd. All rights reserved.