Artificial Neural Networks for Drying Characteristics Determination: An Application in Fluidized Bed Drying


Yıldırım A. , Ergök H., Ağra Ö.

5th International Anatolian Energy Symposium, Trabzon, Turkey, 24 - 26 March 2021, pp.11-20

  • Publication Type: Conference Paper / Full Text
  • City: Trabzon
  • Country: Turkey
  • Page Numbers: pp.11-20

Abstract

Drying have a several usage of purpose which are extend the shelf life of food products, to reduce packaging costs and shipping weights, to protect original flavors and nutritional values. In recent years, with increased interest in organic products, dried vegetables and fruits are in great demand. In this study, the drying characteristics of button mushrooms have been investigated in fluidized bed dryer. Button mushrooms (Agaricus Bisporus) are rich protein source and covers 37,7% of the world's fungal production. Experimental study was carried out in fluidize bed dryer at air temperatures of 45-50-55-60 °C at constant air velocity of 6,5 m/s and at constant 50 °C drying air temperature of 5.5-6.5-7.5 m/s. Mushrooms cut like a slab with thickness of 5 mm and experimental drying times have been examined. Experimental data were used as inputs of artificial neural network and a network was trained. With the help of the trained network, the best experimental conditions were selected by calculating the error rates between the inputs and outputs. Thus, test costs, equipment usage rates and energy consumption were reduced.