LSTM-BASED SOLAR POWER FORECASTING

The world population, which is increasing day by day, causes the energy needs of countries to increase. This increase in energy demand has led especially developing countries to seek different energy sources to use in energy production. Greenhouse gas emissions and harmful gases resulting from energy production, 70% of which is provided by the use of fossil fuels, cause the ozone layer to thicken, which accelerates global warming. In order to find solutions to energy needs and to prevent global warming to some extent, the trend towards more environmentally friendly energy sources is increasing day by day. At the same time, the use of natural resources such as sun and wind for energy production enables countries to become energy independent. However, the variable structure of renewable energy sources compared to traditional energy sources increases the cost of using these resources in energy production. Nowadays, avoiding these disadvantages has become much more economical with developing technology. Nowadays, the ability to collect and store countless data has enabled the study areas of Machine Learning algorithms, which are algorithms that learn by imitating the human brain, to become increasingly widespread. The main motivation of this study is to make the calculations required for Solar Energy Systems installations comprehensive and economical with the help of machine learning algorithms, in order to increase the use of solar energy, which is one of the renewable energy sources. In this context, the solar energy system production data set in the study was used in future time predictions with the help of deep learning models. The predictions made were compared with the help of different statistical calculations and graphs and the model that made the best prediction was discovered.