Application of Genetic algorithm on IC substrate drilling path optimization

This paper proposes genetic algorithm (GA) based heuristic algorithms for the integrated circuit (IC) substrate drilling path optimization problem. IC substrates are cut from a piece of copper clad laminate which contains tens of thousands of holes. A path connecting these holes has to be plotted for the drilling operation. The objective in this research is minimizing the total path length. Thus, this problem is treated as a large scale traveling salesman problem. In this research, we developed three effective algorithms, unit optimization algorithm, two units incremental algorithm, and local fine-tune algorithm. This study employs a numerical experiment to determine the appropriate combination of algorithm parameter values. Test instances obtained from the case company were used to evaluate the performance of the proposed algorithms. It is shown that unit optimization algorithm is able to effectively obtain the approximate optimal value similar to the solutions of case company in fastest time. Furthermore, the two unit incremental algorithm and local fine-tune algorithm successfully improve execution time and solution stability.