Data-Driven Modeling, Monitoring, and Process Optimization of Infrared-Convective Drying for Pumpkin Powder Production

Pumpkin is a valuable source of carotenoids, vitamins, minerals, and dietary fiber and is widely used in the production of functional food ingredients. However, its high moisture content (88 ± 2%, wet basis) limits storage stability and requires efficient drying technologies. This study investigates an infrared-convective (IR-CV) drying process for pumpkin powder production using mathematical modeling and process monitoring approaches. Experimental investigations were conducted using Lechebnaya and Arabatskaya pumpkin varieties at drying temperatures of 60 °C and 70 °C. A mathematical model based on heat and mass balance equations was developed to describe temperature distribution, moisture diffusion, and mass transfer during the drying process. Preliminary infrared treatment was also evaluated to improve drying performance. The results showed that preliminary infrared treatment accelerated moisture removal and improved drying efficiency. After 240 min of drying, the residual moisture content decreased from 23% to 21% at 60 °C and from 21% to 19% at 70 °C. Process optimization indicated that maintaining product temperatures below 65 °C, selecting an infrared wavelength of 2.8 μm, and using a product layer thickness of 4–8 mm provided the most suitable operating conditions. The developed model showed good agreement with experimental data (R² > 0.95) and enabled reliable monitoring of the drying process. The proposed approach can improve process stability, reduce energy consumption, and enhance product quality, demonstrating its potential for intelligent monitoring and future machine learning-based control applications in food processing systems.