Design and Development of Artificial Neural Network-Based Prediction Model for Hemispherical Solar Still

This work reports the prediction of hemispherical solar still water productivity with help of Artificial neural network (ANN). Single slope solar still design has been adopted and modified into hemispherical solar still by converting the single slope solar still basin into circular basin to change into dome shaped solar still. In this study three different spherical cap angles (i.e., 90 (hemispherical), 75 and 60) of dome were taken. Thermal analysis has done to get the glass, water and basin temperatures and water productivity of solar stills. The water productivity of single slope solar still and hemispherical solar still has been compared. From the results, it is found that hemispherical solar still is producing an average of 350 ml more water than single slope solar still under same conditions. The data got from theoretical analysis was used as input to ANN for predicting the water productivity of solar stills. The developed ANN model was predicting the water productivity with accuracy of 98%.