Application of mathematical and genetic algorithm-artificial neural network models in microwave drying of sprouted quinoa

Document Type : Original research

Authors

1 MSc Student, Department of Food Science and Technology, Faculty of Food Industry, Bu-Ali Sina University, Hamedan, Iran

2 Associate Professor, Department of Food Science and Technology, Faculty of Food Industry, Bu-Ali Sina University, Hamedan, Iran

Abstract

Quinoa is one of the pseudocereal grain that is rich in macro- and micronutrients. Sprouting is an effective process that improves the palatability, quality, nutritional value, and digestibility of quinoa seeds. In this research, the impact of microwave dryer power on the drying kinetics and moisture loss of sprouted quinoa was investigated. The sprouted quinoa seeds were dried as single layers at three different power levels (330, 440, and 550 W). The results showed that the drying time was decreased with increasing microwave power. Seven kinetic models were examined to simulate the experimental drying kinetics and the Page model showed the best performance. The effective moisture diffusivity coefficient (Deff) was calculated to be in the range of 1.43×10-10 m2/s to 2.93×10-10 m2/s, and increased significantly with increasing microwave power (p<0.05). The average rehydration ratio of dried sprouted quinoa changed from 251.04% to 290.10%, and increased with increasing microwave power. In addition, in this study a genetic algorithm-artificial neural network (GA-ANN) method was used for prediction of the moisture loss of sprouted quinoa. The optimal network containing four neurons in hidden layer was able to predict the moisture loss of sprouted quinoa with a coefficient of determination (r) of 0.996. The highest values of water loss, water diffusion, and rehydration rate were obtained when drying with a microwave power of 550 W. The results of this study can be useful in selecting optimal drying conditions for microwave drying of sprouted quinoa and as a basis for other sprouted crops.

Keywords

Main Subjects


Articles in Press, Accepted Manuscript
Available Online from 25 April 2025