Research And Application Of Short-term Wind Power Forecasting Based On WRF Model And Cluster Analysis

Wind energy is a kind of clean and pollution-free renewable energy,and it has unlimited potential,inexhaustible,the use of wind power generation is extremely environmentally friendly,so increasingly by the world's attention.However,due to the intermittency and instability of wind power generation and the serious impact of geographical location on the utilization of wind power,large-scale wind power grid connection will have a huge impact on the power grid.Therefore,to ensure the safe and stable operation of the power grid and to admit as much wind power as possible under the conditions of safe and stable operation of the power grid,the research on wind power prediction becomes increasingly important.This paper conducts an in-depth study on short-term wind power prediction,and the main contents are as follows:Firstly,the wind energy of the target wind field is preliminarily predicted by using the numerical weather forecast-based WRF model.According to the actual geographical location of the wind field,an appropriate parameterization scheme of WRF model is configured.The model outputs the predicted wind speed and direction.By comparing the predicted wind speed and direction with the actual observation data of wind field,it is found that the predicted wind speed and direction have good consistency with the actual observation data.However,there are still some errors between the wind speed and wind direction data predicted by WRF and the actual data.In order to improve the prediction accuracy of WRF model,ant colony-RBF neural network algorithm is introduced to correct the predicted wind speed.By comparing the wind speed data before and after the correction,according to the error evaluation index,the wind speed predicted by the revised WRF model is closer to the actual observation data,which proves the effectiveness of the proposed correction method.Secondly,in view of the large number of fans in the wind field and the large and slow workload of modeling a single fan,the fuzzy c-means clustering algorithm is introduced to group and combine the fans in the wind field.Artificial fish-elman neural network algorithm was used as a prediction model to model and predict each fleet.Finally,the predicted value of each cluster is weighted and aggregated to obtain the total actual predicted value of the wind farm.By comparing the predicted value with the actual value,it is found that the method combining clustering algorithm and artificial fish-elman neural network can improve the accuracy of wind power prediction.Lastly,on top of the existing wind power prediction system,the wind speed correction algorithm and power prediction algorithm mentioned above are integrated into the wind power prediction system,and the reliability of the algorithm are verified in the actual system application.