Résumé
Objective To determine the moisture in Huodan pills by near-infrared spectroscopy. Methods At the beginning,water content of 174 samples were determined and all data of samples about near-infrared spectrum were collected. And then all the data were pretreated by first-derivative and Savitzky-Golay filter. At last,the NIR quantitative model for moisture in huodan pills was established by partial least squares regression and validated by the validation set. Results The correlation coefficients,the root mean square error overall of calibration, that of cross validation of the calibration model was 0. 988 17, 0. 103,and 0. 326 16,respectively. The root mean square error overall of prediction and the average recovery of validation was 0. 217 and 98. 02%. Conclusion The NIR spectroscopy analysis model is steady,accurate and reliable,which can be used to detecte moisture content of huodan pills.
Résumé
<p><b>OBJECTIVE</b>To establish a near-infrared spectroscopy quantitative model for rapid determination of the patchouli alcohol content in Pogostemon cablin.</p><p><b>METHOD</b>The gas chromatography was adopted for determining the content of patchouli alcohol content in 102 batches of P. cablin samples. Their near-infrared spectrograms were collected and preprocessed by standard normal variate and the first derivative of Savitsky-Golay. The quantitative model of patchouli alcohol content was established by the partial least squares regression analysis.</p><p><b>RESULT</b>According to the correction model established in this study, the root-mean-square error of calibration, the root-mean-square error of prediction and the root-mean-square error of cross-validation of the calibration model for Patchouli alcohol were 0.991 10, 0.012 9, 0.012 8 and 0.033 15, respectively.</p><p><b>CONCLUSION</b>The near-infrared spectroscopy quantitative model established in this study is stable, accurate and reliable for the rapid determination of the content of patchouli alcohol in P. cablin.</p>