Research Paper
Mohammad Ganje; Masumeh Moghimi; Hamid Bakhshabadi
Abstract
Today, the use of non-thermal methods such as osmotic drying, dehydration using ultrasound, and microwave are increasing for the production of food products. This study aimed to predict the trend of changes in some characteristics of osmotic dried medlar fruit with microwave pretreatment under the influence ...
Read More
Today, the use of non-thermal methods such as osmotic drying, dehydration using ultrasound, and microwave are increasing for the production of food products. This study aimed to predict the trend of changes in some characteristics of osmotic dried medlar fruit with microwave pretreatment under the influence of different microwave times (0 to 4 min), temperatures (40 to 60°C), and osmotic process times (60 to 240 min). To predict the trend of changes in these characteristics, response surface methodology with a central composite design and artificial neural network was utilized. It was determined that the optimal microwave time, temperature, and osmotic process time should be 1.69 min, 60°C, and 240 min, respectively, to minimize the solid gain and to maximize the percentage of water loss and the of water loss/solid gain. Under these conditions, the solid gain was 15.17%, water loss was 24.21%, the water loss/solid gain was 1.63, and weight reduction was 10.00%. Through the examination of different networks, the feedforward back-diffusion network with 4-5-3 topologies was found to be the best neural model, with a correlation coefficient greater than 0.996 and a mean square error less than 0.001. The hyperbolic tangent activation function, hopping learning model, and 1000 learning cycles were also determined to be optimal. Comparing the correlation coefficients of the models obtained from the response surface method and the artificial neural network, it was concluded that the artificial neural network method was more effective in predicting the trend of changes in the characteristics of dried medlar fruit through the osmotic process with microwave pretreatment.
