近红外在线检测技术在人参提取物生产中的应用  

Application of Near-infrared Spectroscopy On-line Detection to Production Process of Ginseng Extract

作　　者：张瑞[1] 王英平[1] 刘宏群[1] 许世泉[1] 闫梅霞[1] 赵景辉[1] 

机构地区：[1]中国农业科学院特产研究所,吉林吉林132109

出　　处：《特产研究》2012年第2期30-33,42,共5页Special Wild Economic Animal and Plant Research

基　　金：科技部十一五科技支撑计划(2007BAI38B03);国家科技部项目(2006BAI06A14)

摘　　要：采用比色法和近红外漫反射、液体透射在线检测技术,分别测定93个不同产地、不同年生人参原材料固体样品和558个人参提取液样品的总皂苷含量和近红外光谱,结合偏最小二乘回归法,建立样品总皂苷含量与近红外光谱之间的数学模型。采用内部交叉检验对模型进行评价。评价结果显示:内部交叉检验的决定系数R2>0.95,交叉检验均方差(RMSECV)值分别为0.071 6和0.032 7,相对分析误差(RPD)分别为2.83和3.97。模型评价数据均有较好的表现,原材料、提取液检验样品近红外测定结果与化学参考值之间无显著性差异,近红外检测技术可以应用于人参原材料固体样品和人参提取液样品中总皂苷含量的快速测定。This experiment aims to use near-infrared spectroscopy to generate the spectrum of Chinese ginseng samples. The total ginseno- sides contents and near-infrared spectroscopy figure were determined in raw materials of ginseng and 558 extract solutions. Raw materials of ginseng came from 93 different places, and born in different years. Using partial least squares （PLS）, the methematical models were developed to correlate near infrared spectra and total ginsenosides contents. Method of cross-validation was adopted to evaluate these models, the results showed： The correlation coefficients of model were all above 0.95, the root mean square error of cross-validation （RMSECV） values were 0.071 6 and 0.032 7, ratio of deviation to performances （RPD） of model finally developed were 2.83 and 3.97 respectively. The model achieved the best prediction performance, and there is no significant difference between predictive contents and chemical reference values. The result indicates that NIR spectroscopy could he used in practice to fast determine the total ginsenoside contents of powder and extract solutions.

关 键 词：近红外光谱 质量检测 过程分析 总皂苷 

分 类 号：Q503[生物学—生物化学]