摘要
采用偏最小二乘法(PLS)建立测定八角茴香中莽草酸含量的近红外(NIR)光谱定量分析模型。应用多种光谱预处理方法分别对八角茴香固体粉末样品的NIR光谱进行预处理,并采用预处理后的光谱建立定量分析模型,每个模型均经过选择最有效的光谱区域和最适主因子数进行优化。经过比较各个模型的内部交互验证均方根误差(RMSECV)和交互验证预测值与真实值间的相关系数(RV),外部预测均方根误差(RMSEP),选取最优的模型,结果表明定量分析模型稳健性好和测定精度高,在中药有效成分定量分析方面有很好的应用前景。
Partial least-square （PLS） method was applied to establish quantitative analysis model using near-infrared （NIR） spectra for determination of shikimic acid content in aniseed （ lUicium verum Hook. f. ）. Various spectrometric pretreatment methods were used respectively for conversion of NIR spectra of aniseed powdered samples in order to remove the noise in spectra. The pretreated spectra were applied to develop PLS quantitative analysis model respectively. Each model was optimized by selecting the most efficacious spectrometric pretreatment method and the most suitable number of factors. The optimum model was selected depending on the root mean squares of calibration set calculated by cross-validation （ RMSECV）, the correlation coefficient of the values prediction by cross-validation method and actual values and the root mean squares of prediction set （RMSEP）. The results demonstrated that the optimum model possessed good stability and high precision. This method has great prospect for quantitative analysis of Chinese traditional medicine.
出处
《林产化学与工业》
EI
CAS
CSCD
2008年第5期26-30,共5页
Chemistry and Industry of Forest Products
基金
中国医学基金会新药发展基金(20061108)
关键词
莽草酸
八角茴香
近红外光谱
偏最小二乘法
shikimic acid
lllicium verum Hook. f
near-infrared spectroscopy
partial least-square method
