人工神经网络-近红外光谱法非破坏监控蛇床子SFE-CO_2萃取物的质量  

Application of Artificial Neural Network-near-infrared Reflectance Spectroscopy in Controlling Extractant of Fructus Cnidii by SFE-CO 2

作　　者：郭晔[1] 王彬[1] 曲楠[1] 赵羚志[1] 李文良[1] 弥宏[1] 任玉林[1] 

机构地区：[1]吉林大学化学学院,长春130012

出　　处：《吉林大学学报（理学版）》2008年第2期341-345,共5页Journal of Jilin University:Science Edition

基　　金：吉林省科技厅中药现代化项目基金(批准号:20030607-2)

摘　　要：将人工神经网络与近红外漫反射光谱相结合,对蛇床子超临界CO2萃取物中的蛇床子素和欧前胡素同时进行快速、非破坏的定量分析.分别利用萃取物的原始光谱和两种预处理光谱(一阶导数和标准归一化)建立了网络的数学校正模型,并把所得结果进行了比较,确定用一阶导数光谱所建立的模型为最佳网络模型.A rapid and non-destructive analytical method for simultaneous analysis of osthol and imperatorin in the extractant of Fructus Cnidii obtained by supercritical carbon dioxide fluid extraction （SFE-CO2 ） was developed by artificial neural network based on near-infrared spectroscopy. The artificial neural network （ANN） models of original spectrum and two pretreated spectra （ first-derivative and standard normal variate, respectively） of osthol and imperatorin were established. In the models the concentrations of osthol and imperatorin were determined simultaneously and compared. The best model of determining the concentrations of osthol and imperatorin was obtained via their first-derivative spectra.

关 键 词：人工神经网络 近红外漫反射光谱 蛇床子 超临界CO2萃取 

分 类 号：O657.33[理学—分析化学]