Research on Ultra-Short-Term Prediction of Wind Power Based on Improved Wavelet BP Neural Network Algorithm

In order to ensure the safety and stable operation of the power system and rationally arrange the maintenance plan of the wind turbine, this paper proposes an improved wavelet BP neural network algorithm for the research of ultra-short-term power forecasting of the wind farm. In order to greatly improve the accuracy of wind power prediction and simultaneously consider the delay characteristic of the prediction model, this paper makes use of the wavelet discrete transform for the frequency band decomposition of the signal, then combines the genetic algorithm to model the BP neural network and then simulates the output signals of each layer. Finally, combined with the simulation results and the calculation method of the wind power forecasting calculation by the National Energy Administration, it can be demonstrated that this improved method can provide theoretical support for the prediction system of wind farm power.