| 基于主成分分析和小波神经网络的近红外多组分建模研究 |
| |
| 引用本文： | 汤守鹏,姚鑫锋,姚霞,田永超,曹卫星,朱艳.基于主成分分析和小波神经网络的近红外多组分建模研究[J].分析化学,2009,37(10). |
| |
| 作者姓名： | 汤守鹏 姚鑫锋 姚霞 田永超 曹卫星 朱艳 |
| |
| 作者单位： | 南京农业大学江苏省信息农业高技术研究重点实验室,南京,210095 |
| |
| 基金项目： | 教育部新世纪优秀人才支持计划，国家自然科学基金，国家科技支撑计划，江苏省创新学者攀登计划，江苏省自然科学基金 |
| |
| 摘 要： | 将小麦叶片原始光谱经过预处理后,采用主成分分析(PCA)对数据进行降维,取前3个主成分输入小波神经网络,建立了基于主成分分析和小波神经网络的近红外多组分预测模型(WNN);进一步研究了小波基函数个数的选取(WNN隐层节点数)对小波神经网络模型性能的影响,并将WNN模型与偏最小二乘法(PLS)和传统的反向传播神经网络(BPNN)模型进行了比较.结果表明,所建立的WNN模型能用于同时预测小麦叶片全氮和可溶性总糖两种组分含量,其预测均方根误差(RMSEP)分别为0.101%和0.089%,预测相关系数(R)分别为0.980和0.967.另外,在收敛速度和预测精度上,WNN模型明显优于BPNN和PLS模型,从而为将小波神经网络用于近红外光谱的多组分定量分析奠定了基础.
|
| 关 键 词： | 小波神经网络 主成分分析 近红外光谱 小麦叶片 全氮 可溶性总糖 |
Near Infrared Multi-component Prediction Model Based on Principal Component Analysis and Wavelet Neural Network |
| |
| TANG Shou-Peng,YAO Xin-Feng,YAO Xia,TIAN Yong-Chao,CAO Wei-Xing,ZHU Yan.Near Infrared Multi-component Prediction Model Based on Principal Component Analysis and Wavelet Neural Network[J].Chinese Journal of Analytical Chemistry,2009,37(10). |
| |
| Authors: | TANG Shou-Peng YAO Xin-Feng YAO Xia TIAN Yong-Chao CAO Wei-Xing ZHU Yan |
| |
| Institution: | TANG Shou-Peng,YAO Xin-Feng,YAO Xia,TIAN Yong-Chao,CAO Wei-Xing,ZHU Yan(Jiangsu Key Laboratory for Information Agriculture,Nanjing Agricultural University,Nanjing 210095) |
| |
| Abstract: | A new near infrared model for multi-component prediction was established based on principal component analysis(PCA) and wavelet neural network(WNN) methods.First,original near infrared spectra from wheat leaves were pre-processed,and their principal components were extracted by PCA,which could reduce the dimensionality of spectrum data.The first three principal components were taken as inputs of wavelet neural network(WNN),and the influence of neuron number in the hidden layer of WNN on the properties of mo... |
| |
| Keywords: | Wavelet neural network principal component analysis near infrared spectrum wheat leaf nitrogen content soluble sugar content |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
