摘要
近红外光谱技术具有简便、快速和无损检测等优点,应用可见-近红外光谱方法建立生姜水分含量(moisture content)的预测模型。利用可见-近红外光谱仪采集308个生姜的光谱,其光谱范围是350～1 800 nm。分别采用一阶导数(FD)、二阶导数(SD)、标准正交变量变换(SNV)和多元散射校正(MSC)4种方法对光谱进行预处理,结合偏最小二乘法(PLS),分别在430～1 000 nm、1 000～1 800 nm、430～1 800 nm 3个波段建立生姜水分含量的PLS预测模型。对实验结果进行分析表明,在波段范围430～1 800 nm使用一阶导数预处理方法建立的PLS模型最优。其验证组的相关系数为0.975 1,预测组的相关系数为0.959 7。结果表明,可见-近红外光谱可以准确、快速地对生姜的含水量进行检测。
Near-infrared reflectance spectroscopy has advantages of rapid determination,non-destruction,and convenient.In this work,visible-near infrared spectrum was used to establish the prediction model for ginger moisture content.Visual Near-infrared spectroscopy（Vis-NIRS） combined with partial least squares（PLS） was used for the prediction of moisture content in the ginger.The spectra of 308 ginger samples were collected with a wavelength range of 350~1 800 nm.First derivative FD,second derivative SD,standard normal variate SNV and multiplicative scattering correction MSC were adopted for the acquisition of Vis-NIR spectra.In the three bands with ranges of 430~1 000 nm,1 000~1 800 nm and 430~1 800 nm the PLS model were established respectively.This study shows that the PLS model established in the range of 430~1 800 nm,using first derivative spectra pretreatment method,was the optimal.The validation group of the correlation coefficient and that of the forecasting group were 0.975 1 and 0.959 7 respectively.The results showed that Visual Near-infrared spectroscopy（Vis-NIRS） can detect the moisture content of ginger accurately and rapidly. Cr and Pb reached the lower pollution level,Cu and Ni pollution had different pollution levels in the studied area.The ecological risk index method indicated that about 80% of the studied area reached severe and very severe harm degree and Cd was the main element of potential harm,the follower was As in the studied area.
出处
《江西农业大学学报》
CAS
CSCD
北大核心
2011年第3期602-607,共6页
Acta Agriculturae Universitatis Jiangxiensis
基金
国家自然科学基金资助项目(30760101)
江西省教育厅科学技术研究项目(GJJ08513)
关键词
可见-近红外光谱
生姜
水分含量
无损检测
偏最小二乘法
surface sediments
heavy metal
distribution characteristics
pollution assessment
Zhushan Bay of Taihu Lake
