Solar Power Generation Prediction Using Machine Learning and Study of Power Generation Using Solar Tracking Panels

As the importance of renewable energy is growing worldwide, Republic of Korea aims to increase the proportion of renewable energy generation to 30.2% by 2030. Solar power plays an important role, accounting for more than 80% of domestic renewable energy generation. However, since the amount of solar power generation varies greatly depending on weather factors, it is necessary to respond to the instability of solar power generation and increase the generation amount in order to secure stable energy. Therefore, in this paper, we learn a machine learning model based on weather data to predict the amount of solar power generation, calculate the solar altitude and azimuth that change according to seasons and time, and increase the amount of solar power generation by controlling the solar power panel. As a result of the experiment, the power generation was predicted to be within 0.208 mean square error, and it was confirmed that the difference in power generation before and after sun tracking was 23.4%.