BP neural network model based on cluster analysis for wind power prediction

As wind has the property of intermittency and randomness, it is very important for the wind plant to improve the accuracy for wind power prediction to satisfy the stability requirements when combined to the grid. This paper studies the major factors that influence wind power, and proposes a BP neural network model based on cluster analysis with the traditional BP neural network model for comparison. The results show that the new model can significantly improve the wind power prediction accuracy. © 2011 IEEE.