Abstract
This research deals with the real time scheduling problem in the photolithography area of semiconductor wafer fabrications. The objective is to find an optimal schedule that minimizes the weighted function of makespan, maximum tardiness, and setup frequency. Practical issues like machine breakdowns, limited number of masks, restrictions of photo-resist, production restriction, and machine setups are considered in this study. A heuristic algorithm (TPI rule) to generate initial solution and three kinds of search algorithms (tabu search, simulated annealing, and genetic algorithm) are applied to solve the scheduling problem. Moreover, we propose a new approach of sensitivity search. Whenever a new event occurs that change the original scheduling problem, instead of searching from a random or a heuristic initial solution, we restart the search from the best solution of the previous problem, which is very similar to the original problem. Computational experiment is conducted to test this approach. The results show that the performance of the sensitivity search significantly surpasses the traditional approach.