Akademik Veri Yönetim Sistemi
15th International Congress on Agricultural Mechanization and Energy in Agriculture – ANKAgEng’2023, Antalya, Türkiye, 29 Ekim - 01 Kasım 2023, cilt.1, ss.44
In this research, nonlinear models including artificial neural network (ANN) and adaptive-network-based fuzzy inference system (ANFIS) were employed to evaluate the moisture ratio and drying rate of the Rheum ribes L. dried by infrared/convective dryer. The drying process was carried out in three temperatures (50, 60 and 70 °C), infrared powers (250, 500 and 750 W), the distances between the product and the infrared source (10, 15 and 20 cm), and thicknesses (3, 5 and 7 mm). These four parameters along with the drying time (0-1000 min) were considered as the network input in the ANN and ANFIS to assess the moisture ratio and drying rate. The best nonlinear model was selected based on the determination coefficient (R2), mean square error (MSE), and mean absolute error (MAE). The rise in the temperature and IR power and a decline in the thickness and the distance between the product and the infrared source can shorten the drying time. The R2 value of ANN and ANFIS models was 0.9988 and 0.9996, and MSE values was 0.0037 and 0.0021, respectively. Based on the considered statistical criteria (R2, MSE, and MAE), the best model for the prediction of melon moisture content was the ANFIS model. These results indicated the high performance of the ANFIS model in the evaluation of the moisture content compared to ANN methods.