OBJECTIVE:To establish a quantitative model for the online analysis of hesperidin content in the extraction process of Huatan Qushi decoction for individuals with phlegm-dampness constitution by using near-infrared (NIR) spectroscopy and partial least squares (PLS) methods.METHODS:The reference content of hesperidin,an effective ingredient of Huatan Qushi decoction,was determined using high-performance liquid chromatography (HPLC).The modeling band was selected using synergy interval PLS (SiPLS) algorithms,and a PLS model demonstrating the relationship between NIR predictive values and HPLC reference values was established.The optimal modeling route was selected through global optimization of process trace parameters.RESULTS:The root mean square error of cross-validation was 9.8410.The coefficients of determination of the calibration set and prediction set were 0.9860 (R2cal) and 0.9458 (R2pre),respectively.The residual predictive deviation was 3.0115.The parameters demonstrated high predictive capability and reliability.The band was selected via the SiPLS method.There were seven latent variables.CONCLUSION:This study provides a favorable comprehensive evaluation method for variable selection and quality control of the Traditional Chinese Medicine extraction process.