Short-Term Wind Power Forecast for Wind Farm Base on Artificial Neural Network

Wind power forecasting is a critical method in minimizing the impact caused by integrated power grids. Firstly, the general steps of the establishment of neural network based load forecasting model for wind farm are presented and the principles are introduced. Next, in an example of a wind farm, the paper focuses on the relationship of weather data and output power by analyzing this relationship. A number of factors having significant impact on the output power are selected and used as network input. This network is an excellent predictive model which has been proved high prediction accuracy in experiments.