Solar power production forecast based on weather data in Finland

Abstract

Solar power forecasting has become an important factor in Europe in the recent past, particularly in the middle Europe as well as in the Nordic countries such as Denmark and Finland. The need for accurate forecasting has played a pivotal role in planning the operations of photovoltaic (PV) systems as well as in achieving power grid balance. In this thesis, a statistical model for solar power forecasting is computed, studied, investigated and used to predict solar power. The model uses past power measurements and meteorological forecasts of temperature, solar irradiation, relative humidity and wind speed as inputs. The weather forecast parameters used to compute power are obtained from Aladin Research Model on Nonhydrostatic forecast Inside Europe (HARMONIE) representing Lappeenranta region. The computed estimate power is then compared with the real power produced from Lappeenranta University of Technology (LUT) solar power plant. Normalized Root Mean Square Error (NRMSE) is used as the evaluation criteria. The results indicate that solar power production can be forecasted using the model with small NRMSE errors captured indicating better performance of the model.