Abstract: Correlation analyses of antioxidant properties were used for selecting an appropriate method for quality evaluation of dried loquat. Arrhenius first order kinetics and difference models were established with drying time and temperature as independent variables versus the contents of flavonoids and total phenols and 2,2’-azino-bis(3-ethylbenzothiazoline-6- sulphonic acid) (ABTS) free radical scavenging capacity. The models were tested by mean error, mean absolute error and root mean square error. Nonlinear programming and evolutionary algorithms were used to minimize the energy consumption. Results showed that flaovonoids, total phenols and antioxidant activity demonstrated a slow upward trend with the loss of water during the drying process of loquat. The Arrhenius difference models improved the accuracy of predictions compared with the first order kinetics models, with determination coefficients of 0.92, 0.90 and 0.93 for flavonoids, total phenols and ABTS free radical scavenging capacity, respectively. Mean error, mean absolute error and root mean square error of validation were between 0.07–0.80, 0.18–1.53 and 0.07–0.59, respectively. The models were found to be valid in terms of precision. Optimal drying time and temperature were (183 ± 1) min and (65 ± 1) ℃, respectively.

Key words: loquat, drying, antioxidant activity, difference model, process optimization