An Artificial Intelligence Wind Power Prediction Model Based on Meteorological Parameters

Abstract

In the light of the problems of large fluctuation of wind power, complex influencing factors, and low accuracy of traditional single forecasting model. This paper investigates the correlation of the meteorological parameters and wind power, and combines the different characteristic of BiGRU, CNN, and attention mechanism, the hybrid CNN-BiGRU-Attention based wind power prediction model has been proposed. The model can provide more accurate wind power forecasts for wind farms. In addition, the experimental conclusions of this work illustrate that the forecast accuracy of the model which uses wind speed and wind power as inputs can superior to other models, with higher prediction accuracy and stability.