Artificial intelligence predictive thermodynamic analysis and energy enhancement of conveyor belt dryer under combined heating conditions

Mathematical modeling and artificial intelligence play a critical role in optimizing food drying processes. This study innovatively combines thermodynamic, kinetic, and energy performance modeling of an IR-HA conveyor-belt dryer for Moringa oleifera leaves with multi-criteria optimization utilizing Artificial Neural Networks (ANN) and Self-Organizing Maps (SOM). Drying experiments were conducted at air velocities ranging from 0.3 to 1.0 m/s, air temperatures of 30–50 °C, and infrared intensities of 100–600 W/cm². It was observed that increasing the air temperature from 30 to 50 °C and the IR intensity from 100 to 600 W/cm² significantly shortened the drying time by up to 42 % and reduced the specific energy consumption (SEC) from 4.3 to 2.1 MJ/kg. In contrast, higher airflow increased the drying time by 18 % and raised the SEC. The drying process exhibited an accelerating followed by a falling rate period, with the Midilli and Kucuk model providing the best fit to experimental curves (R² > 0.99, RMSE < 0.004). Maximum drying and thermal efficiencies of 29 % and 27 %, respectively, were achieved. To enhance predictive capability, Artificial Neural Networks (ANNs) and Self-Organizing Maps (SOM) were integrated with thermodynamic and kinetic modeling, enabling accurate prediction and clustering of optimal operating conditions. The proposed framework offers practical guidance for scaling up IR–HA conveyor drying of temperature-sensitive products, ensuring energy efficiency, sustainability, and quality preservation. This investigation examines the limitations of previous research on energy optimization and efficiency analysis of Moringa leaves using a conveyor belt dryer. Additional studies are necessary to assess the economic feasibility of implementing such a system for sample drying and to understand how drying affects the thermal properties of the products.