Computational intelligence and machine learning Approaches for performance evaluation of an infrared dryer: Quality analysis, drying kinetics, and thermal performance

Abstract

The quality of agricultural products is frequently compromised by energy-intensive drying methods, highlighting the need to develop more efficient drying techniques. Infrared drying has surfaced as a compelling approach for enhancing energy efficiency and product quality. This study examines the thermal characteristics, physicochemical properties, and drying kinetics of garlic slices subjected to different infrared radiation intensities (0.1–0.3 W/cm²), air temperatures (35–55 °C), and airflow rates (0.5–2 m/s). An artificial neural network (ANN) with 99 % predicting accuracy was employed to refine drying parameters, aiming to improve drying efficiency and minimize specific energy consumption (SEC). The findings demonstrate that elevated air temperature and infrared intensity significantly decreased drying time, with the minimum duration of 290 min recorded at 55 °C, 0.3 W/cm², and 0.5 m/s airflow. The SEC was reduced to 3.78 MJ/kg under these optimal conditions. The increase in infrared radiation and temperatures resulted in a decrease in allicin content, dropping from 17 % to 11.53 %, as well as a reduction in vitamin C retention, which fell from 0.112 mg/g to 0.05 mg/g. Nonetheless, there was an enhancement in thermal efficiency, achieving a peak of 51.9 % at 55 °C and 0.1 W/cm². The ANN model exhibited impressive predictive accuracy in estimating drying time, SEC, and thermal efficiency. The results offer significant insights for enhancing infrared drying technology, presenting a sustainable method to decrease energy usage while maintaining key quality characteristics of dried garlic.