Meteorological Data Driven Prediction of Global Solar Radiation

Abstract

This paper aims at prediction of global solar radiation (GSR) using artificial neural network (ANN) model. Various meteorological parameters such as relative humidity, average air temperature, wind speed, wind direction, dew point and atmospheric pressure contribute to estimation of global solar radiation. A search has been carried out to find the strongly correlated variables which contribute more to this assessment. On that basis, an ANN model is devised for the prediction of GSR. The results are compared with those obtained using multiple linear regression (MLR) model. The performances of the models are checked using various statistical parameters such as mean absolute deviation (MAD), mean square error (MSE), root mean square error (RMSE), mean absolute percentage error (MAPE) and correlation coefficient (R). A comparison of the predicted values obtained from the two methods and the actual values obtained from a weather monitoring station shows that the developed ANN model generates values closer to actual. The best MAPE calculated by using ANN model is 2.851%.