Solar Power Prediction in Different Forecasting Horizons Using Machine Learning and Time Series Techniques

Abstract

Solar power generation is highly intermittent, nonlinear, and variable in nature. The increase in penetration level of solar energy resources poses technical challenges. An accurate forecasting model is crucial to minimizing these technical issues. Therefore, choosing the right forecasting technique for the right forecasting horizon is vital. In this study, the performance analysis of machine learning and time series forecasting techniques for various forecasting horizons has been investigated. Its accuracy, root mean square error (RMSE), and mean absolute error (MAE) have been compared to other techniques.