FORECASTING POWER GENERATION OF SOLAR PANELS USING MACHINE LEARNING

Abstract

Solar energy, a renewable resource, is becoming an increasingly important part of the global energy source. This energy, which is known for its clean and abundant amount, has become the subject of numerous research studies. Advances in science and technology have made the use of solar energy more common and widely used. However, the operation of the solar system sometimes encounters difficulties due to constant weather fluctuations, which significantly affect the output capacity of the photovoltaic solar panels. This can lead to an electricity surplus when the load consumes less power or a lack of power supply when there is insufficient energy to produce it. To solve this problem, this research focuses on using Machine Learning methods to predict the output capacity of solar panels based on weather data. There are two main data sources used in the research: the data provided and the actual data collected through the measurement process achieved. The input data for the prediction model includes ambient temperature, wind speed, humidity, and total solar radiation from the provided dataset, as well as ambient temperature and solar illuminance from the collected dataset. The output of the prediction model is the power generation from solar panels. The results of the study will help the power system operate more efficiently, ensuring the balance between supply and demand, thus utilizing the potential of solar energy effectively.