Comparative Analysis of Machine Learning Algorithms to Predict Solar Irradiance

In recent times, there has been a rapid shift from non-renewable to renewable sources of energy. Recently there has been a lot of development in photovoltaic systems that use solar irradiance energy as a source. Solar energy is a clean source of energy that can be tapped to meet the demands of the load/grid. Prediction of the amount of solar irradiance at a location can be beneficial for optimum production. This study compares the accuracy of several machine learning algorithms, including SVM and LSTM, for the prediction of solar irradiance at specified latitudinal and longitudinal coordinates using historic data from NASA POWER. It is observed that solar irradiance is influenced by factors like temperature, precipitation, humidity among others.