Seyed Mostafa Hosseinalipour - Assistant Professor Faculty of Mechanical Engineering, Iran Univ. of Science & Tech.
Mehdi Bahiraei - Ph.D. Student Faculty of Mechanical Engineering, Iran Univ. of Science & Tech.
Ehsan Taheran - MSc Student Faculty of Mechanical Engineering, Iran Univ. of Science & Tech.
Payman Taherkhani - MSc Student Faculty of Mechanical Engineering, Iran Univ. of Science & Tech.

In this study, an experimental impinging jet oven equipped with an online data acquisition system was developed. A software pack was 'Mitten for data acquisition. The baking experiences were carried out at oven temperatures of 150, 175, 200, 225 and 250°C and five different jet velocities. Then, 1025 series of acquired data were employed in Neural network (NN) training and testing process. This NN was able to predict bread moisture and thickness variations as a function of bread temperature, baking time and jet temperature and velocity. The best NN structure, regarding the minimum difference between experimental and NN outputs, consisted of one hidden layer with 12 neurons which predicts bread moisture and thickness variations reasonably good for the period of baking. R2 of NN outputs in comparison with experimental results were obtained 0.969 for moisture and 0.958 for thickness.

Impinging Jets, Experimental, Iranian Bread, Measurement, Neural Network