论文部分内容阅读
目的:应用近红外光谱技术(NIR)建立快速测定天南星药材中水分的方法。方法:采用烘干法测定样品中的水分,运用偏最小二乘法(PLS)建立该含量与NIR光谱之间的多元校正模型,采用相关系数(R),校正均方根误差(RMSEC),内部交叉验证均方根误差(RMSECV)和外部预测均方根误差(RMSEP)对校正模型进行优化和评价。利用校正模型对未知样品的水分进行预测,检验模型的准确度。结果:采用二阶导数法对光谱进行预处理,在4 774~9 845 cm~(-1)波段,选择前6个主成分建立最优校正模型,所建模型的R为0.990 6,RMSEC为0.16,RMSECV为0.38。经外部验证,校正模型的RMSEP和平均回收率分别为0.298和99.8%。结论:该方法具有简便快速、结果准确、无损样品的特点,可以应用于天南星中水分的快速测定。
Objective: To establish a rapid method for determination of water in Araceae by near infrared spectroscopy (NIR). Methods: The water content in samples was determined by drying method. The multivariate calibration model was established by partial least squares (PLS) between the content and NIR spectra. Correlation coefficient (R), root mean square error (RMSEC) The RMSECV and RMSEP were used to optimize and evaluate the calibration model. The calibration model is used to predict the moisture content of unknown samples to test the accuracy of the model. Results: The spectra were preprocessed by the second derivative method. The optimal calibration model was established by selecting the first six principal components in the band of 4 774 ~ 9 845 cm -1. The R of the model was 0.990 6, RMSEC was 0.16, RMSECV is 0.38. External validation showed that the RMSEP and the average recovery of the calibration model were 0.298 and 99.8%, respectively. Conclusion: The method is simple, rapid, accurate and non-destructive. It can be applied to the rapid determination of water content in Araceae.