黄花蒿中青蒿素的近红外光谱法测定

Determination of Arteannuin in Artemisia annua L. by Near Infrared Spectroscopy

采用近红外光谱数据及其小波变换处理后的小波系数,以偏最小二乘法建立模型并预测预示集中黄花蒿样品的青蒿素含量。结果表明:小波变换充分提取了近红外光谱的信息,数据压缩为原始数据量的3.3%。利用C5小波系数对青蒿样本进行建模并预测,RMSEP和r2分别为0.544‰和0.988。

The model was established by PLS method with the NIR spectra and wavelet coefficient from wavelet transform. Arteannuin in Artemisia annua L. could be predicted by this model. The results indicated that the wavelet trarisform could extract the information of NIR spectra, the data were compressed to 3.3 %. The RMSEP and r^2 of the validation set samples of the model base on the C^5 wavelet coefficient were 0.544‰ and 0.988, respectively.