Özet:

Turkey’s annual production value is 750,000 tons, a fruit with a high nutritional and economic value. The drying process recognizes the possibility of extending the lifetime of the roof as well as the possibility of producing a dry roof, which is a product preferred by the consumer. The capacity to model the processes of drying the cave by advanced methods can greatly help product processors make production planning and cost analysis. The artificial nerve network (YSA), a machine learning method that has found quite a wide scope in recent years, is used to solve complex problems in various tasks, from parameters and function predictions to classification. This study aims to model the traditional (with hot air), electrohydrodynamic (EHD) and EHD-heat air combination with the drying methods of the different drying parameters of the dried pot and the effects on the drying speed and product moisture with the YSA. By testing different transfer functions and learning algorithms, YSA has been identified to provide the best model performance for each drying method. The determination ratio on the test data of the YSA models for the EHD-heat air combination and forecast of drought and moisture in the coating drying was determined as values higher than 0.96. Research findings have shown that the EHD method and agricultural product drying can be modeled by YSA-based methods.

Anahtar Kelimeler: