Experimental and theoretical investigation of shelled corn drying in a microwave-assisted fluidized bed dryer using Artificial Neural Network

Drying characteristics of shelled corn (Zea mays L) with an initial moisture content of 26% dry basis (db) was studied in a fluidized bed dryer assisted by microwave heating. Four air temperatures (30, 40, 50 and 60 °C) and five microwave powers (180, 360, 540, 720 and 900 W) were studied. Several experiments were conducted to obtain data for sample moisture content versus drying time. The results showed that increasing the drying air temperature resulted in up to 5% decrease in drying time while in the microwave-assisted fluidized bed system, the drying time decreased dramatically up to 50% at a given and corresponding drying air temperature at each microwave energy level. As a result, addition of microwave energy to the fluidized bed drying is recommended to enhance the drying rate of shelled corn. Furthermore, in the present study, the application of Artificial Neural Network (ANN) for predicting the drying time (output parameter for ANN modeling) was investigated. Microwave power, drying air temperature and grain moisture content were considered as input parameters for the model. An ANN model with 170 neurons was selected for studying the influence of transfer functions and training algorithms. The results revealed that a network with the Tansig (hyperbolic tangent sigmoid) transfer function and trainrp (Resilient back propagation) back propagation algorithm made the most accurate predictions for the shelled corn drying system. The effects of uncertainties in output experimental data and ANN prediction values on root mean square error (RMSE) were studied by introducing small random errors within a range of ±5%.