用神经元网络预测靶材与薄膜性能间的关系

Investigation on Relationship between Target and Properties of Thin Film Using Neural Network

用神经元网络的方法预测了溅射靶材、工艺参数与电阻薄膜性能之间的关系．经预测误差分析以及与实验比较，表明用７９组训练样本可基本准确预报电阻薄膜的性能．对各参数影响薄膜性能的程度也作了分析．

A preliminary investigation using neural network is performed on the relationship between target composition, processing and properties of thin film resistor. The neural network is trained using 79 model vectors. It is shown that, by error analysis and comparison with experimental data, the neural network is able to predict the properties of thin film resistor fairly well. Furthermore, the sensitivity analysis of processing variables on the properties of thin film is made based on the prediction from neural network.