Solar Irradiance Prediction using Deep Learning-Based Approaches

The worst effects of pollution from greenhouse gas emissions are increase in atmospheric temperature, sea level rise, and a variety of diseases. The main source of greenhouse gases is the combustion of fossil fuels, especially in thermal power plants. The use of renewable energy sources, such as solar energy, can significantly reduce this pollution. The amount of electricity produced from solar energy can be maximised by placing the solar power plant in an area with high direct horizontal irradiance (DHI). Solar irradiance predictions will help to design and build solar power facilities very efficiently. However, accurate prediction is a challenge. In this study, the deep learning approaches such as single-layered LSTM and stacked LSTM variations like two-layered LSTM and three-layered LSTM are used to forecast DHI.