Vacuum drying of sweet cherry: Artificial neural networks approach in process optimization

Abstract

Sweet cherries were vacuum-dried and the process was optimized with relative antioxidant activity index (RACI), standard score (SS), and artificial neural network (ANN) approach. Investigated input parameters were drying temperature and pressure, while moisture content, water activity, total phenolic, flavonoid and anthocyanins content, and antioxidant activity (FRAP, DPPH, and ABTS test) were analyzed as output parameters. The obtained mean value of R2 (0.872) indicates that the ANN investigated in this research could be successfully applied for describing the sweet cherry vacuum drying in the range of temperatures from 50 to 70°C and of pressures from 20 to 200 mbar. The optimized sweet cherries vacuum-dried process presents significant support for the possibility toward application of vacuum drying for sweet cherries in industrial conditions. According to ANN results, it is possible to take any combination of input parameters and calculate the output parameters observed in this research. Practical applications In the framework of this research, the artificial neural network (ANN) were applied on sweet cherries vacuum drying and as a main result the optimized vacuum drying process was obtained. Such optimized process presents significant base for the possibility toward application of sweet cherries vacuum drying in industrial conditions. This is enabled, since according to ANN results obtained in this research it is possible to take any set of input vacuum drying parameters (T: 50–70°C and p: 20–200 mbar) and calculate the values of the output parameters (moisture content, water activity, total phenolic, flavonoid and monomeric anthocyanin content, and antioxidant activity) observed in this research.