Forecasting Day-Ahead Solar Radiation Using Machine Learning Approach

Abstract

Unpredictability of solar resource poses difficulties in grid management as solar diffusion rates rise continuously. One of the big challenges with integrating renewables into the grid is that their power generation is intermittent and unruly. Thus, the task of solar power forecasting becomes vital to ensure grid constancy and to enable an optimal unit commitment and cost-effective dispatch. Latest techniques and approaches arise worldwide each year to progress accuracy of models with the vital aim of reducing uncertainty in the predictions. This paper appears with the aim of compiling a big part of the knowledge about solar power forecasting, focusing on the most recent advancements and future trends. Firstly, the inspiration to achieve an accurate forecast is presented with the analysis of the economic implications it may have. To address the problem and we rummage superlative prediction models for forecasting solar radiation using machine learning techniques. We compare multiple regression techniques for generating prediction models, including linear least squares and support vector machines using multiple kernel functions. In our experiments, we analyze predictions for day ahead solar radiation data and show that a machine learning approach yields feasible results for short-term solar power prediction. A root mean square error improvement of around 32% is achieved by the proposed model compared to others proposed reference model except one.