Performance evaluation of a solar hybrid collector with dryer experimental and machine learning approach

This study investigates the drying efficiency of mint leaves using three methods—open sun drying, natural convection, and forced convection in a solar hybrid collector with a dryer—achieving average drying efficiencies of 12.4%, 18.7%, and 24.5%, respectively. Additionally, machine learning models—support vector regression (SVR) and random forest (RF)—were employed to evaluate the drying process. Across all three methods, a significant reduction in mass was observed. The electrical and thermal performance of the solar hybrid collector with a dryer was also analyzed on the basis of sustainable practices. The mass loss of mint leaves during natural and forced convection drying was approximately 51% and 82%, respectively. Drying experiments were conducted using four aluminum trays arranged from bottom to top (Tray 1 to Tray 4) within the solar hybrid collector. The highest moisture removal rate of 78.23% was recorded in Tray 1, with a decreasing trend in Trays 2, 3, and 4, achieving 76.48%, 75.32%, and 72.13%, respectively. Among the three drying methods, forced convection proved to be the most effective, while open sun drying was the least efficient. The predictive performance of RF and SVR models was evaluated, yielding R² values of 0.8723 and 0.8611, respectively, with RMSE values of 0.8125 and 0.7845.