| Title |
A Study on Prediction of Solar Power Generation based on Synoptic Meteorological Observation Data |
| Authors |
김준용(Jun-Yong Kim) ; 정재원(Jai-Won Chung) |
| DOI |
https://doi.org/10.5573/ieie.2023.60.4.93 |
| Keywords |
Solar power; Linear regression; Synoptic meteorological data; Photovoltaic power; Weather data |
| Abstract |
Recently, most of the studies on prediction of solar power generation are conducted to predict the amount of power generation by using machine learning methods or meteorological data to overcome the uncertain characteristics of power generation output of new and renewable energy. However, machine learning models require a relatively large amount of computation for construction and operation, and as a result, there is a high possibility that errors for actual data will increase due to overfitting. In this paper, we propose a solar power generation prediction model using multiple linear regression analysis and meteorological information closely related to power generation. The proposed prediction model collected solar power generation data and meteorological data collected for 149 hours and analyzed the correlation between each meteorological variable and power generation. Afterwards, multiple linear regression analysis was performed on each independent variable, and a multiple linear regression model was constructed to predict power generation by selecting only statistically significant variables. To evaluate the performance of the implemented multiple linear regression model, the coefficient of determination, RMSE, and MAE were calculated, respectively, and the performance was compared with solar power generation prediction models using existing machine learning methods. |