| 多组分相关分析技术在近红外检测中的应用研究 |
| |
| 引用本文： | 陈斌,蔡斌.多组分相关分析技术在近红外检测中的应用研究[J].安徽农业科学,2008,36(17):7077-7078. |
| |
| 作者姓名： | 陈斌 蔡斌 |
| |
| 作者单位： | 江苏大学食品与生物工程学院,江苏镇江,212013;江苏大学食品与生物工程学院,江苏镇江,212013 |
| |
| 基金项目： | 浙江省科技计划重点项目(2004C21043). |
| |
| 摘 要： | 采用信号分析理论中的相关分析技术,将三组分的涤纶-棉-毛样品与二组分的棉-毛样品进行相关分析,提取出三组分样品中的棉-毛信息,并对毛的含量进行定量分析,在31个样品中选22个样品建立了基于偏最小二乘(PLS)方法的NIR(Near Infrared,NIR)校正模型,对余下的9个样品进行了预测分析。结果表明,借助相关分析算法,毛含量的预测相对均方差和相关系数分别从3.5575、0.9855提高到2.7835、0.9876。试验证明,多组分相关分析技术可以在复杂混合样品中提取出待测组分的近红外光谱信息,为建立较高精度的校正模型创造了条件。
|
| 关 键 词： | N1R 校正模型 偏最小二乘 相关分析 |
| 文章编号： | 0517-6611(2008)17-07077-02 |
| 修稿时间： | 2008年4月7日 |
Application of Correlation Analysis of Multi-component Technology in the Near-infrared Detection |
| |
| CHEN Bin.Application of Correlation Analysis of Multi-component Technology in the Near-infrared Detection[J].Journal of Anhui Agricultural Sciences,2008,36(17):7077-7078. |
| |
| Authors: | CHEN Bin |
| |
| Abstract: | With the theory of signal analysis of the related technology,correlation analysis of the three components of polyester-cotton-wool samples and the two components of cotton-wool were conducted.The cotton-wool information of three components samples were extracted and quantitative analysis of wool was studied.22 samples were chosen in 31 samples to Constitute NIR(Near Infrared,NIR) correction model based on the partial least squares(PLS)method.And the remaining 9 samples were forecasted.The results showed that the predicted relative mean square deviation and correlation coefficients rose from 3.557 5 and 0.985 5 to 2.783 5 and 0.987 6,respectively,by using correlation analysis.Test proved that the correlation analysis of multi-component technology could extract the NIR spectral information in complex mixed samples,which created condition for the establishment of higher precision correction model. |
| |
| Keywords: | NIR Correction model Partial least squares Correlation analysis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
