A Comparative Study of Machine Learning Algorithms for Photovoltaic Degradation Rate Prediction

Abstract

Solar energy is the most versatile, harmless, and non-exhaustive energy present in nature because of this, the number of photovoltaic modules that have been integrated into the electrical grid is increasing every day. As a result, reliable forecasting of falling power output over the period of time is required for an acceptable return on investment made for these interactions, to estimate the power delivered to the power system by these photovoltaic modules, photovoltaic degradation rates must be known. In this study degradation rates of photovoltaic modules are estimated using the application of nine machine learning models and the effectiveness of these models is compared in order to determine which model is most efficient. All the models are tested on various evaluation metrics like mean absolute error, root mean squared error, and mean percentage error for an unbiased evaluation, and the run time of these models is also calculated and compared to determine the overall efficiency of the models.