Prediction of Solar Radiation Using Machine Learning Algorithms

Abstract

Design of solar energy systems and its installation at any location critically depends on the Global Solar Radiation (GSR) available at that location. However, it is not practical to do GSR measurements at every location because of the instruments that will be used. This paper discusses the study, modelling and prediction of global solar radiation (GSR) using machine learning (ML) algorithms. The objective of this paper is to use the meteorological parameters like GSR, temperature and wind speed measured at a location and develop a mathematical model using Machine Learning Algorithms. Such models can be used to predict GSR when direct measurement is not possible. In this study, various Machine Learning algorithms such as Linear Regression, Random Forest Regression, Artificial Neural Networks (ANN) and Deep Neural Networks (DNN) are used for modelling. Among these models, the most suitable method is selected based on the Root Mean Squared Error (RMSE) and Coefficient of Determinant (R2) values. It is found from the study that prediction of GSR with good accuracy is possible using Machine Learning algorithms.