Prediction of Photovoltaic Power Generation using Machine Learning - A Review

The world environment crisis; leading towards zero carbon emission on one hand and the increase in electrical energy demand on the other hand has accelerated us to make better use of the non-conventional energy sources present on earth. Sun being the major source of non-conventional energy produces clean energy however, the weather conditions, day-night patterns, and seasons affect renewable energy production a lot. Machine learning can be a powerful tool to rescue us from this uncertainty in the case of renewable power production. This paper is a comprehensive review of various machine-learning techniques for predicting solar power generation by keeping track of solar irradiance, temperature, and other parameters that affect solar power generation. The paper provides insight into the ML techniques used previously, their benefits, and the best-suited multi-level ML techniques for the prediction of solar power generation. Further, we wind up our work by concluding the future scope of the discussion and proposing the ML methodologies that can be employed in the future in the field of power generation prediction using machine learning.