摘要
建立了氮化锆薄膜制备工艺参数与薄膜色度参数之间的人工神经网络预测模型,结果表明,预测结果与实测结果吻合,最大色差在5.45以内。利用所建立的模型研究了单个参数对薄膜颜色的影响规律,及多参数间交互作用与薄膜颜色的关系。并且利用神经网络根据加工要求反向预测工艺参数,从而实现了对加工参数的优化选择。
A chromatic prediction model was developed by artificial neural network through analyzing the relationship between the technological parameters for ZrN thin film preparation and relevant chromatic parameters. As a result, the prediction is proved basically in conformity to the measurement that shows a max. chromatism within 5.45. The model was used to investigate how a single technological parameter and the interactions between different parameters affect the film color. According to the requirements during film preparation, a backward prediction was also done introducing the artificial neural network, thus optimizing the selection of technological parameters.
出处
《真空》
CAS
北大核心
2007年第2期37-39,共3页
Vacuum
关键词
氮化锆薄膜
颜色
人工神经网络
预测
ZrN thin film
chromatics
artificial neural network
prediction
