Multi-objective optimization of convective drying of apple cubes

Abstract

The effect of drying temperature and air velocity on apple quality parameters, such as color difference (CD), volume ratio (VR) and water absorption capacity (WAC) in convective drying was experimentally studied. Optimization of drying conditions was carried out in the range of air temperatures from 50 to 70 °C and air velocity from 0.01 to 6 m s−1. A novel algorithm of multi-objective optimization, based on artificial neural network (ANN), genetic algorithm (GA) and Pareto optimization was developed. Three optimization objectives included simultaneous minimization of CD, maximization of VR and maximization of WAC. Objective functions for CD, VR and WAC were developed by using ANN training on the experimental dataset of apple drying at 50, 60 and 70 °C. Pareto optimal set was developed with elitist non-dominated sorting genetic algorithm (NSGA II). Unique Pareto optimal solution within specified constraints was found at air temperature 65 °C and velocity 1 m s−1. This mode of apple drying resulted in CD = 5.24, VR = 49.66% and WAC = 0.488. Experimental verification showed that maximum error of modelling did not exceed 3.24%.