Pizhūhish va Nuāvarī dar ̒Ulūm va Sanāyi̒-i Ghaz̠āyī (May 2012)
Estimation of tomato drying parameters using artificial neural networks
Abstract
In this research we have simulated drying tomato thin layer by hot air convection. Tomato slices were dried in two temperatures 60° and 70℃. Perceptron neural network was used to predict moisture ratio and drying rate of samples during the drying process. Best neural network topology for ANN-I based on one hidden layer 2 and 8 neuron per hidden layers for moisturizing ratio and the drying rate obtained respectively. Furthermore, best neural network topology for ANN-II based on one hidden layer 11 neuron for moisturizing ratio and the drying arte obtained. Generally, the results showed that ANN-II had preferable result to predict drying parameters of drying tomato.
Keywords