首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于独立分量和神经网络的近红外多组分分析方法
引用本文:方利民,林敏.基于独立分量和神经网络的近红外多组分分析方法[J].分析化学,2008,36(6):815-818.
作者姓名:方利民  林敏
作者单位:中国计量学院计量技术工程学院,杭州,310018
摘    要:采用小波变换对光谱数据进行压缩,用独立分量分析(ICA)方法提取近红外光谱数据矩阵的独立成分和相应的混合矩阵,再用BP神经网络对混合矩阵和实测浓度矩阵进行建模,提出了基于独立分量分析-神经网络回归(ICA-NNR)的近红外分析建模方法。进一步研究了独立分量数和网络中间隐层的神经元数对模型性能的影响,经优化后的ICA-NNR模型在相关系数与均方根误差两个指标上均优于直接用光谱矩阵作为输入所建立的模型。本方法用于玉米中水分、淀粉、蛋白质3种主要成分含量的同时测定,检验样品集的化学检测值与近红外预测值的相关系数分别达到:淀粉r=0.971,蛋白质r=0.976,水分r=0.975。

关 键 词:独立分量分析  神经网络  小波变换  近红外光谱  玉米样品

A Method of Near Infrared Multi-component Analysis Based on Independent Component Analysis and Neural Networks Model
FANG Li-Min,LIN Min.A Method of Near Infrared Multi-component Analysis Based on Independent Component Analysis and Neural Networks Model[J].Chinese Journal of Analytical Chemistry,2008,36(6):815-818.
Authors:FANG Li-Min  LIN Min
Abstract:A new method of model construction based on wavelet transform(WT),back-propagation artificial neural networks(BP-ANN)regression and independent component analysis(ICA)was proposed.In its application to near infrared spectrum,the data of NIR spectrum were firstly compressed by wavelet transform,their independent components and the mixing matrix were then extracted by the independent component analysis,and finally,the model of the maxing matrix and concentration matrix was built by the artificial neural networks regression.The influence of the numbers of independent components and the neurons in the hidden layer on the properties of model was further analysed.Compared to the model made by direct use of spectrum matrix as input of the neural network,ICA-NNR(neural network regression)model has advantages in both the correlation coefficient and root mean square error indicators.This new chemometric method has been applied to the simultaneous determination of three components(moisture,protein,starch)of corn samples.The correlation coefficients(r)between the chemically tested values and the near-infrared method predicted values of starch,protein and moisture are 0.971,0.976 and 0.975 respectively.
Keywords:Independent component analysis  neural networks  wavelet transform  near infrared spectrum  corn
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号