Scheduling of pressing process in multi-layer printed circuit board manufacturing via milp and heuristic

Abstract

The pressing process aims to press the panel which is the stack of materials to form a multi-layer printed circuit board (PCB). This process is a part of multi-layer PCB fabrication and can be considered as a scheduling problem with the objective of minimizing the makespan. In this dissertation, two mixed-integer linear programming models (Models 1 and 2) and a three-phase-PCB-pressing heuristic (3P-PCB-PH) algorithm for scheduling the pressing process are presented. Model 2 is an improvement of Model 1 in terms of the model size and the dimensionality of some decision variables. Both models and the 3P-PCB-PH algorithm are used to solve the test problems that are generated from the actual data from a PCB company. The results show that Model 2 can find an optimal solution in more test problems than Model 1 and outperform Model 1 in terms of the size complexity and the computational complexity with 34.71% average relative improvement of the computational time. Moreover, the 3P-PCB-PH algorithm can solve all test problems and give an optimal solution or a near optimal solution using the computational time of less than 1 second, which is very practical in the real PCB manufacturing industry.