论文部分内容阅读
本文用裂解质谱方法,得到漆器漆膜裂解质谱数据,然后用变步长误差反向传播算法的人工神经网络模型处理,初步结果证明人工神经网络方法为解析古代漆膜的年代信息提供了一条有效的新途径.本文同时还用聚类分析法研究了漆膜裂解质谱数据隐含的年代信息,并讨论了实验条件的稳定性对解析结果的影响.
In this paper, cleavage mass spectrometry method was used to obtain lacquer film cleavage mass spectrometry data, and then processed by artificial neural network model with variable step error back propagation algorithm. The preliminary results proved that artificial neural network method provided an effective method for analyzing the age information of ancient lacquer film This paper also uses the cluster analysis method to study the implied chronological information of the film lysis mass spectrometry data and discusses the influence of the stability of experimental conditions on the analytical results.