Abstract: The integration of large scale wind power into grid will pose a threat to power system operation in terms of security and economy, so it is necessary to predict wind power. The wind power prediction approach based on CFD pre-calculated flow fields solves the contradiction between forecasting precision and efficiency skillfully, which has high engineering practicability. In order to further improve the wind power prediction method based on CFD pre-calculated flow fields, a one-year wind power is predicted in an actual wind farm, and the error distributions of predicted results are studied. The forecasting results show that the prediction approach has high accuracy and can well predict various ways of power generation, which meets the application requirements of wind power prediction. The error frequency distributions change with seasons, and forecasting power has some transverse or longitudinal errors. In addition, different wind speed and direction sectors are correspond to different forecasting errors. The inhomogeneous wind conditions discrete methods can be adopted to predict the wind power, and then the efficient and reasonable application of the prediction approach can be realized.