水平衰减全反射-傅里叶变换红外光谱结合化学计量学法应用于三种缩叶藓属植物的分类  

Classification of Three Kinds of Ptychomitrium by Horizontal Attenuated Total Reflection-FTIR Combined with Chemometric Methods

作　　者：孔黎春[1] 余鹏[1] 程存归[1] 

机构地区：[1]浙江师范大学化学系,金华321004

出　　处：《理化检验（化学分册）》2012年第4期405-409,共5页Physical Testing and Chemical Analysis(Part B:Chemical Analysis)

基　　金：浙江省公益技术应用研究计划项目(分析测试2011C37051)

摘　　要：采用水平衰减全反射(HATR)-傅里叶变换红外光谱法(FTIR)测定了3种缩叶藓属植物齿边缩叶藓、多枝缩叶藓和中华缩叶藓的红外谱图,运用离散小波变换对吸收较为相似的3种缩叶藓属植物的红外谱图进行特征提取。通过分析比较后选择第三,四分解层进行特征向量的提取,利用所得到的特征变量进行径向基神经网络(RBF-NN)训练,再将训练出来的网络对不同产地的3种缩叶藓属植物的红外谱图离散小波提取后的特征向量进行分类。通过对120个不同样本的验证,说明能够采用基于FTIR-离散小波进行数据压缩后进行特征变量的提取及径向基神经网络分类法对3种缩叶属植物齿边缩叶藓、多枝缩叶藓和中华缩叶藓进行分类。The horizontal attenuated total reflectance Fourier transform infrared spectroscopy （HATR-FTIR） was applied to classification of 3 ptychomitrium, including ptychomitrium dentatum （Mitt.） Jaeg. , Ptychornitriurn polyphylloides （C. Muell. ） Par and Ptychomitrium sinense （Mitt.） Jaeg. The technique of discrete wavelet transform was used to extract the features of the similar FTIR of the 3 Ptychomitrium. Through analyzing and comparing the characteristic vector extraction in 3, 4 detail decomposition layers and training the feature variables via radial basis function neural network （RBF-NN）, the characteristic extraction vectors of the KTIR of the 3 Ptychomitrium were classified by the network which training out by RBF-NN. By means of the verification of 120 different samples, it was shown that the feasibility of establishing the models with FTIR-DWT-RBF-NN method to identify among Ptychomitrium dentatum （Mitt.） Jaeg. , Ptychornitrium polyphylloides （C. Muell. ） Par. and Ptychomitriurn sinense （Mitt.） Jaeg.

关 键 词：水平衰减全反射-傅里叶变换红外光谱法 离散小波特征提取 径向基神经网络 缩叶藓属植物 分类 

分 类 号：O657.33[理学—分析化学]