RESUMO
Objective: To establish a real-time moisture monitoring model for the fluidized bed drying process of Guizhi Fuling Capsules (GFC) by using online near-infrared spectroscopy (NIRS). Methods: A total of 176 samples from 16 production batches were collected by NIRS diffuse reflection probe for modeling. The moving window average smoothing method was used for spectral preprocessing. The characteristic variables were 4 759.45—5 338.00 cm−1, 5 503.84—6 101.67 cm−1, and 8 512.25—8 809.24 cm−1, which were screened by the interval partial least squares method (siPLS) combined with the moving window partial least squares (mwPLS). The partial variable least squares (PLS) method was used to build a multivariate correction model for moisture. Results: The root mean square error of cross-validation (RMSECV) of predicted moisture was 0.243%, the ratio of predicton to deviation (RPD) was 13.384, and the relative standard error of prediction (RSEP ) was 0.270%. The reliability of the online monitoring method was continuously verified by eight production batches. The relative error of 40 samples was less than 4.7%, indicating that the PLS quantitative model prediction performance was robust and accurate. The real-time monitoring trend chart of the moisture in the drying process can accurately determine the drying end point, and the moisture content of the end sample was within the control limit. Conclusion: The quantitative model established by online NIRS combined with PLS can be applied to the on-line monitoring of moisture content in the fluidized bed drying process of production scale GFC and the prediction performance was robust and accurate.
RESUMO
This study was aimed to optimize the near infrared (NIR) variable selection method based on multivariate detection limit (MDL). Using Qing-Kai-Ling (QKL) injection as object, three variable selection methods (interval par-tial least-squares, iPLS; backward interval partial least squares, BiPLS; moving window interval partial least squares, mwPLS) were used to establish the PLS models of baicalin in QKL injection, respectively. The prediction ability of different variable selection method was compared. MDL of all models were calculated in contrast to the MDL value of full spectra PLS model, to select optimal variable selection method. The results showed that different variable selec-tion methods had different prediction ability. Among them, iPLS had the best performance which determination coef-ficient of prediction (Rpre2) and the root mean square errors of prediction (SEP) were 0.996 5 and 602.3 μg·mL-1, re-spectively. All MDLs of different variable selection methods were reduced compared with the full spectra PLS model. The value of iPLS was the lowest comes to be 1.19 μg·mL-1. The results above indicated that the best variable se-lection method for baicalin in QKL injection was iPLS. MDL theory took the error of calibration and validation set and the leverage of external sample into account, which can comprehensively evaluate model detection performance compared to the classic chemical indicator parameters. This method was particularly suitable for the variable selec-tion method optimization of NIR quantitative model of low concentration sample such as Chinese herbal medicine.