Artificial neural network modeling in predicting thermal performance and assessment of quality parameters during...

Artificial neural network (ANN) models have become essential for precise predictions and improving engineering systems. This study investigated the effects of air velocity (2.0, 3.0, and 4.0 m s^-1), grain flow rate (5.5, 7.0, and 8.5 kg min^-1), and infrared temperature (40, 50, and 60 °C) on drying kinetics, thermal performance, and quality properties in corn using a graphene-based far-infrared dryer. An ANN was used to predict optimal drying conditions to balance heating characteristics for corn quality attributes.

The results showed that a 4.0 m s^-1 air velocity, 60 °C infrared temperature, and 8.5 kg min^-1 flow rate reduced drying time from 8.5 to 3.5 h, lipase activity from 18.92% to 10.78%, and acidity content from 1.88 to 1.22 g NaOH kg^-1. The highest drying conditions achieved the lowest energy usage, resulting in a maximum thermal efficiency of 82.28% at minimum drying time. However, increasing the infrared temperature to 40-60 °C while maintaining the same 5.5 kg min^-1 grain flow rate and 4.0 m s^-1 air velocity resulted in an improved antioxidant activity, from 10.22 to 12.11 g catechin gallate equivalents kg^-1 dry weight. The study used precise ANN modeling to highlight the complex interactions between drying parameters and thermal performance, which recorded a strong predictive performance of 99% accuracy. Principal component analysis showed that acidity and energy consumption have commonalities.

This study highlights the potential of cutting-edge computational tools to enhance energy efficiency through graphene-based heating without compromising product quality. © 2026 Society of Chemical Industry.