Purpose

Due to the large number of chemical components in Chinese patent medicine, the quality control analysis is faced with greater challenges. Although different detection techniques, analytical conditions, and sample pretreatments are often employed respectively to quantify the multiple target components in complex mixtures, it is difficult to meet the practical requirement of productive enterprises.

Methods

Tchebichef moments (TMs) were calculated from the raw HPLC–DAD spectra of samples and the linear quantitative models were established by stepwise regression for the eight active components (four alkaloids, one statin, and three glycosides) in Daizongfang (one kind of Chinese patent medicines), respectively.

Results

The correlation coefficients of prediction (\({R}_{p}^{2}\)) were higher than 0.9159, and the spiked recovery rates were from 93.4 to 108.2%. Compared with the N-PLS and MCR-ALS models, the TM models are more accurate and reliable.

Conclusions

This study not only indicates that chemometrics approach has outstanding advantages for the identification of complex signal, but also provides another way for the rapid analysis of quality control in production.