论文部分内容阅读
用模式识别和人工神经网络法总结MOCVD外延生长GaInAsSb薄膜的生长条件与外延层组成的关系。结果表明,气相中TMIn和TMSb的含量、Ⅴ/Ⅲ比和生长温度是影响外延层组成的主要因素。用这些参数作特征变量,以外延层铟含量是否大于06、锑含量是否大于04作为分类标准,可得到良好的模式识别分类效果和人工神经网络交叉检验结果。
The relationship between growth conditions and epitaxial layer composition of GaInAsSb thin films grown by MOCVD epitaxy was summarized by pattern recognition and artificial neural network. The results show that the content of TMIn and TMSb, V / Ⅲ ratio and growth temperature in the gas phase are the main factors affecting the composition of the epitaxial layer. Using these parameters as feature variables, we can get good results of pattern recognition classification and artificial neural network cross-test by whether the content of indium in the epitaxial layer is greater than 06 and the content of antimony is greater than 04.