The optimization of line cycle time in printed circuit board assembly

Abstract

This research addresses the problem of Printed Circuit Board (PCB) assembly in an electronics manufacturing system. In the electronic industry, an assembly line normally has several non-identical component placement machines and the placement times of the different machines for the same component are different. Faced with the global competition, an efficient component placement operation is essential. This research attempts to determine the allocation of components to placement machines for the optimization of the line cycle time. A mathematical model was constructed to represent the mechanism for determining the optimal line cycle time. Initially the mathematical model was in a non-linear integer minimax type formulation. It was then converted into an integer linear programming format. The Branch-and-Bound (B&B) algorithm was applied to solve the integer linear programming model in this research project. However, the B&B alogrithm was found to have taken a very long time to get the optimal solution and hence a heuristic method, the Tabu Search (TS) heuristic was proposed to solve the problem. The performances of both the B&B algorithm and the TS heuristic were compared. The result showed that the TS heuristic can achieve an acceptable solution with a shorter computational time and less number of iterations while the B&B algorithm can guarantee to arrive at the optimal solution. Both the B&B algorithm and the Tabu Search procedure are found applicable to determine the optimal line cycle time in PCB assembly efficiently. Moreover, the cycle time of PCB assembly can be reduced as well as the cost of production by the use of the model and methods presented in the project. A further study is required in order to implement the model and the algorithms developed in this project in a real industrial situation, such as, a graphical user interface.