摘要
应用近红外光谱技术(NIR)和OPUS数据分析软件,对市售黄芩饮片中黄芩苷的含量进行了快速测定.采用偏最小二乘法(PLS)对黄芩饮片的结果与NIR建立校正模型,最小最大归一化法为黄芩苷的最优预处理建模方法,优化校正模型中真实值与预测值之间的相关系数(R2)为0.896 1,内部验证均方差(RMSECV)为0.836,最佳主因子数为9,预测平均偏差为0.92%.NIR具有非破坏性、无污染、重现性好等优点,可以用于市售黄芩饮片中黄芩苷含量的快速测定.
It was used to determine rapidly the content of baicalin of scutellaria baicalensis slices from markets by near infrared spectroscopy（NIR）and the OPUS analysis software .To build the model about the content of baicalin and NIR by partial least square （PLS） .The best modeling method pretreatment was minimum maximum normalization method .The R2 be-tween the true and prediction was 0 .896 1 ,the root mean square errors of cross-validation （RMSECV）was 0 .836 ,the best dimension was 9 and the prediction average deviation was 0 . 92% .NIR ow ns many remarkable advantages that can not be displayed by traditional analysis methods ,such as nondestruction ,no pollution and good reappearance .It can be used rapidly and correctly for determination of content of baicalin of scutellaria baicalensis slices from markets .
出处
《陕西科技大学学报（自然科学版）》
2014年第2期121-124,共4页
Journal of Shaanxi University of Science & Technology
基金
陕西省教育厅自然科学专项项目(12JK1031)
陕西科技大学自然科学基金项目(ZX10-20)
