Performance evaluation using artificial neural network technique and exergetic impact of cold plasma pretreatment on hybrid Ultrasound/Convective drying of ginger slices

In this study, exergy destruction was employed as a key thermodynamic metric, and an artificial neural network (ANN) model was developed to provide a comprehensive assessment and optimization of a hybrid drying process. In the specific case of ginger slices, drying experiments were conducted at three air temperatures (45, 55, and 65 °C) and two ultrasound powers (0.04 and 0.08 kW), while samples were pretreated using cold plasma (CP) for 50 and 100 s prior to drying. Compared with untreated samples dried at higher ultrasound power, the combination of shorter CP pretreatment and lower ultrasound power slightly increased drying time but significantly enhanced exergetic performance, reducing exergy destruction by up to 53.69 %. The ANN model accurately predicted the influence of drying parameters with a coefficient of determination (R 2 ) of up to 0.999 and successfully identified the optimal drying conditions (65 °C, 0.04 kW, and 50 s CP) with lower environmental impact.