A wind power prediction method considering wind turbine operating status based on LSTM

With the growing demand for clean energy, wind power generation has attracted increasing attention. However, traditional wind power prediction methods which rely on statistical models and physical equations have certain limitations, rendering them unable to handle nonlinear and non-stationary scenarios. To address this challenge, this paper proposes a novel wind turbine power prediction method based on the long short-term memory (LSTM) neural networks. The proposed method takes into account not only the historical power and wind speed data, but also the operational state parameters of the wind turbine are introduced as input, thereby enabling a more comprehensive consideration of the effect of the operational state on power generation. By applying the long short-term memory (LSTM) neural networks, the proposed method achieves high prediction accuracy and robustness, demonstrating its promising application prospects in wind turbine power prediction.