Research on Quantitative Prediction of Total Ash in Pollen Typhae by NIRS

Near infrared spectroscopy is used for establishing a rapid and non-destructive analytical method to determine the quantitative of Total Ash in Pollen Typhae, a traditional chinese crude drug. Partial least squares regression (PLSR) was used for building the quantitative prediction model of Pollen Typhae different calibration methods were compared with spectral pretreated methods, wave-number selection and factors of NIRS were discussed in detail. The first derivative and multiplicative scatter correction method was used to predict total ash to correlation coefficients (R), root mean square error of calibration (RMESECV) residual prediction deviation (RPD) and offset obtained by PLSR model were 0.9994, 0.293, 29.4 and 0.013 respectively. The standard deviation was less than 8.2% by the total ash prediction model in the 8 batches of Pollen Typhae from 6% to 27.3%. The predictive result shows the method of this research is rapid, non-destructive and credible, which can be applied to the rapid predict the total ash of Pollen Typhae.