In current work, the response surface methodology using Box-Behnken design was initially used to optimize the fermentation conditions of green tea via consumption rate of tea polyphenol (CROTP) as a response value. The CROTP reached its maximum (25.78 %) when the fermentation pH, ultrasonic time and ultrasonic frequency were set at 5, 15 min and 50 kHz, respectively. Based on the optimization, a novel strategy was further proposed to dynamically monitor the quality change of green tea during ultrasonic-assisted fermentation (UAF) utilizing a colorimetric sensor array (CSA) comprising nine color-sensitive dyes combined with chemometric methods. The fabricated CSA showed sensitive differential responses to green tea fermentation samples, reaching the best quantitative prediction of tea polyphenol content using support vector machines (SVM) with Rc = 0.886, RMSEC = 0.042 mg/g, Rp = 0.862, and RMSEP = 0.043 mg/g. As a result, our work proved that integrating CSA with SVM is viable...

You do not currently have access to this content.