Improving wind power forecasting accuracy through bias correction of wind speed predictions

Abstract

Accurate wind power forecasting is essential for the efficient integration of renewable energy into electricity markets. This study examines the impact of wind speed bias correction on wind power forecast accuracy using various statistical methods. Analysing data from 75 wind turbines across 10 wind farms in Greece, we find that a 1% reduction in wind speed error leads to an average increase of 0.6% in wind power forecast accuracy. Our analysis further reveals that while more sophisticated models generally yield better results, improving total accuracy by about 12%, simpler methods offer comparable accuracy with lower computational costs. Nevertheless, the absolute accuracy achieved by the bias correction methods depends strongly on the initial quality of wind speed forecasts. Therefore, our findings emphasise the importance of preprocessing techniques and high-quality meteorological data in wind power forecasting.