Scheduling And Simulation In Semiconductor Wafer Fabrication

As the most complicated manufacturing system recognized world-wide, the semiconductor wafer fabrication system features a large number of tool groups, long production cycles and reentrance and does not fall into either of the traditional Flow-shop or Job-shop system. The modeling and optimization of the semiconductor wafer fabrication system has attracted huge attention from researchers and scientists in both academic and practical arenas, for this age sees increasingly heated competition in the high-investment, high-profit and high-risk semiconductor manufacturing industry. However, the traditional optimizing methods can not tackle those NP-Hard problems. Thus modeling and simulation have emerged as an important way out. On the basis of discrete events simulation algorithm, this thesis sets up a FAB model on the base of SEMATCH database, and develops a simulating program for the FAB releasing and dispatching systems with the help of object-oriented C++. The program result shows: the improvement in the releasing system brings about greater overall benefits for the whole FAB system than that in the dispatching system; the dispatching system based on the Ant Colony Optimization principles leads to better performance than that based on FIFO principles. The program also simulates the stable and high-quality production of the FAB system when a new wafer is introduced to it.