ANN Modeling of White Mulberry Drying by Microwave Oven

This paper is concerned with developinganartificial neural network (ANN)model for predicting the moisture content, moisture ratio, drying rate and energy efficiency of microwave drying of white mulberry. The drying experiments were carried out at 200, 300, 400 and 500 W of microwave power.Time and power of drying considered as the input variables to the topology of neural network.The results revealed that a network withthe exponential and logistic transfer function and BFGS algorithm with 8 neuron in hidden layer, made the most accurate predictions for the white mulberry drying system. The R2 of best model for training and testing of data were 0.98 and 0.96, respectively. Simulation test on the prediction of network output parameters performed and the accuracy of the predicted values are excellent.

Artificial neural network, BFGS, Microwave drying, White Mulberry