基于遗传算法的显影速率参数确定  

Determination of Development Rate Parameters Based on Genetic Algorithm

作　　者：郝慧娟[1] 张玉林[1] 魏强[1] 卢文娟[1] 

机构地区：[1]山东大学控制学院电子束研究所,济南250061

出　　处：《微细加工技术》2007年第3期6-8,17,共4页Microfabrication Technology

基　　金：国家自然科学基金重大研究计划资助项目(90307003);山东省自然科学基金资助项目(Y2003G03);国家自然科学基金资助项目(10572078)

摘　　要：为了更精确地确定显影速率参数,采用基于遗传算法的最小二乘法对显影速率进行拟合,并与Gauss-Newton迭代法进行了比较。结果显示,基于遗传算法的拟合效果更佳,具有较强的鲁棒性,可以用于显影速率参数的确定。该拟合可以为显影模拟、曝光条件的优化提供更精确的参数。The development rate was simulated using a least square method based on the genetic algorithm to accurately determine the parameters for development rate and it was compared with one by the Gauss-New ton iterative method. The results indicate the simulated result by the genetic algorithm method is more accurate than one by the Gauss-Newton iterative method and it has the stronger robust, so it can be used to determine the development rate parameters. This work can provide the precise parameters for development simulation and the optimization of exposure condition.

关 键 词：显影速率 最小二乘法 遗传算法 Gauss-Newton迭代法 

分 类 号：TN305.7[电子电信—物理电子学]