Predicting the cumulative productivity of a solar distillation system augmented with a tilted absorber panel using machine learning models

Solar distillation plays a crucial role in addressing water purification challenges, making it a key technology in sustainable solutions. To enhance the performance of conventional solar distillers (CSD), this study focused on incorporating an absorber panel as an innovative approach. Two solar distillers were designed, manufactured, and subjected to a 10-hour experimental evaluation, measuring variables such as water temperature, glass covering temperature, ambient temperature, and cumulative productivity. The introduction of the absorber plate resulted in a remarkable increase in productivity, with the modified solar distiller (MSD) achieving a 138.68% improvement, from 1311.3 ml/m2.h to 3129.8 ml/m2.h. The adoption of machine learning techniques for forecasting the accumulated productivity of solar distillation systems holds immense importance in enabling precise and efficient predictions rather than long experimental evaluations. To predict cumulative productivity values, three machine learning models were tested, namely, Support Vector Machine (SVM), Decision Tree (DT) and k Nearest Neighbor (kNN). The kNN algorithm exhibited exceptional performance in forecasting cumulative productivity for both conventional and modified solar distillers, demonstrating a determination coefficient of 1.000 and a zero valued coefficient of variation. These findings highlight the promising potential of machine learning in future research endeavors aimed at forecasting solar distiller outputs.