Multipin cold plasma electric discharge on hydration properties of kodo millet flour: Modelling and optimization using response surface methodology and artificial neural network – Genetic algorithm

Abstract

The effect on functional properties of kodo millet flour was studied using multipin cold plasma electric reactor. The analysis was carried out at various levels of voltage (10–20 kV) and treatment time (10–30 min) for four different parameters such as water absorption capacity (WAC), oil absorption capacity (OAC), solubility index (SI) and swelling capacity (SC). Response surface methodology (RSM) and artificial neural network – genetic algorithm (ANN – GA) were adopted for modelling and optimization of process variables. The optimized values obtained from RSM were 20 kV and 17.9 min. On the contrary, 17.5 kV and 23.3 min were the optimized values obtained from ANN – GA. The RSM optimal values of WAC, OAC, SI and SC were 1.51 g/g, 1.40 g/g, 0.06 g/g and 3.68 g/g whereas optimized ANN – GA values were 1.51 g/g, 1.50 g/g, 0.06 g/g and 4.39 g/g, respectively. Infrared spectra, peak temperature, diffractograms and micrographs of both optimized values were analyzed and showed significant differences. ANN showed a higher value of R² and lesser values of other statistical parameters compared to RSM. Therefore, ANN – GA was treated as the best model for optimization and modelling of cold plasma treated kodo millet flour. Hence, the ANN – GA optimized values of cold plasma treated flour could be utilized for practical applications in food processing industries.