Prediction Method of Wind speed and Wind power Under the Influence of Multi-factor Coupling Under Extreme Weather Conditions

Wind power is greatly influenced by natural factors, and extreme weather will significantly affect wind power output. A new prediction method is proposed to address the issue of reduced prediction performance caused by the limited number of wind power samples under extreme weather conditions, which cannot fully consider the impact of meteorological factors on the dynamic coupling of wind power. Firstly, the TimeGAN algorithm is used to enrich meteorological and power samples, and on this basis, self attention is introduced to capture the auto-correlation of high-dimensional wind power sequences; The cross attention is used to reconstruct the wind power and meteorological factors, mine the temporal and spatial correlations of multidimensional feature sequences, and then extract the corresponding time series characteristics using the Long short-term memory network. The obtained time series characteristics are input into the full connection layer after being fused by the global attention denoising and gating mechanism for wind power prediction. Finally, simulation analysis was conducted, and compared to conventional prediction modes, the proposed method demonstrated good prediction performance in extreme weather conditions.