Artificial Intelligence-Assisted Thermal Energy Conversion and Adaptive Control of a Cabinet-Type Solar Dryer for Medicinal Plant Processing

Solar drying is a sustainable technology that converts solar radiation into thermal energy for processing agricultural and medicinal products. However, variations in solar irradiance, ambient temperature, and relative humidity create nonlinear thermal dynamics that reduce energy conversion efficiency and destabilize drying conditions. Improving thermal energy utilization under fluctuating environmental conditions remains a significant challenge in renewable-energy-based drying systems. This study proposes an artificial intelligence-assisted framework for enhancing thermal energy conversion and adaptive control in a cabinet-type solar dryer. Experimental data were collected from an outdoor drying system equipped with sensors measuring solar irradiance, ambient temperature, relative humidity, and drying chamber temperature. A feed-forward artificial neural network with three input variables (solar irradiance, ambient temperature, and relative humidity), one hidden layer containing ten neurons, and one output variable (drying chamber temperature) was developed using the Levenberg-Marquardt training algorithm. The trained model was integrated into an adaptive predictive control structure to regulate the drying process under changing weather conditions. Model performance was evaluated using the root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). Validation results demonstrated RMSE = 1.15 °C, MAE = 0.87 °C, and R² = 0.985, indicating high prediction accuracy. Compared with conventional passive solar drying operation, the proposed approach reduced temperature fluctuations by 28%, improved thermal energy utilization efficiency by 16%, and shortened drying time by 18%. Predicted temperature profiles closely matched experimental measurements, confirming the reliability of the developed model. The proposed framework transforms a conventional solar dryer into an intelligent renewable-energy conversion system capable of maintaining stable thermal conditions under environmental uncertainty. The results demonstrate the potential of artificial intelligence to improve energy efficiency and adaptive control in solar thermal processing technologies.