Analysis and prediction of incoming wind speed for turbines in complex wind farm: Accounting for meteorological factors and spatiotemporal characteristics of wind farm

Predicting and calculating the incoming wind speed ahead of the turbine hub is a crucial aspect of research into wind power forecasting. This paper proposes a method for predicting wind turbine incoming wind speeds, which considers the meteorological spatial environment, the temporal characteristics of wind speeds, and the effects of topography and wind turbine wake. Firstly, the Wind Meteorological Mast (WMM) wind speed is predicted using the meteorological spatial downscaling and temporal feature extraction methods, which establishes a spatial and temporal relationship between the mesoscale meteorological background and wind speeds at WMM. Secondly, the incoming wind turbine speed is calculated using the WMM-predicted wind speeds, along with topography and wake effects from the WMM to the specific wind turbine are taken into consideration. Thirdly, the performance of the method proposed in this paper was validated using LiDAR for a special wind turbine at a wind farm in Zhangbei, China, and the resulting experimental findings have been subjected to comprehensive analysis. Results indicate that the method presented in this paper can accurately predict the actual incoming wind speed in front of the wind turbine. The hourly single-step incoming wind speed predictions for the subsequent four days indicate that the discrepancies between the actual and predicted incoming wind speed of the MAE, RMSE, R2, and MAPE are 0.6173 m/s, 0.7958 m/s, 0.9432, and 8.466 %, respectively. The incoming wind speed predict method presented in this paper can serve as a reference for wind power prediction.