Short-Term Wind Power Prediction Model Based on WRF-RF Model

In order to improve the accuracy of wind farm output forecasting, a short-term wind power forecasting model combining WRF (Weather Research and Forecasting Model) and random forest (RF) is proposed. This paper takes Qingdao Woer Wind Farm as the research object. The wind speed measurements were performed at four heights (10, 50, 70 and 80 m) of one wind tower attached to the wind farm, and the Pearson correlation coefficients were calculated between the measured wind speed data and the total output power of the wind farm at four different heights to determine the most significant measured wind speed heights associated with the output power of the wind farm. Meanwhile, the WRF is used to predict the meteorological elements, such as wind speed, wind direction, temperature, relative humidity and atmospheric pressure at 70 m. Then, the RF model between the meteorological elements at 70 m and the wind farm output is constructed, the numerical forecast wind power product is refined, the mapping relationship between the two is established, and the wind power output is predicted. The experimental results show that the accuracy of the output power predicted by the random forest model increases as the wind speed increases, and the average absolute error rate is 28.75% when the wind speed is greater than 7m/s. Wind speed has a great influence on wind power prediction, and improving the accuracy of wind speed prediction is an important way to improve wind power prediction.