Application of genetic algorithm optimization on the physical parameters of the meteorological forecast deposition of thin film semiconductor manufacturing process

• Main Authors: ChengHsienLi, 李政憲
• Other Authors: 楊棠堯
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/20273463781622884917

碩士 === 南台科技大學 === 資訊管理系 === 97 === TFT-LCD (Thin Film Transfer Liquid Crystal Display) is a complex process of high-tech industries, its manufacturing processes are highly dynamic process of degeneration or variability. There are many factors such as wear of machinery, materials consumption, and seasonal factors will affect the process parameters to result in producing bad products. In order to control effectively the TFT manufacturing process parameters, this study uses genetic algorithms to forecast these process parameters to construct controls in Physical Vapor Deposition (PVD) process. Multi-variable system for the PVD process, the application of genetic algorithms to deal with polynomial algorithm in order to look forward to reduce the equipment exists in the manufacturing process of the potential impact of variability, and shortened the time to solve this polynomial. This study uses he actual PVD equipments in the TFT industry as our simulation subjects. This study uses genetic algorithms to learn critical parameters through the computer simulation, and to reduce the frequency of manual adjustment, to save costs and reduce the experimental verification of the time.