Research On Wind Power Prediction Based On Spatial Correlation

With the continuous growth of the installed capacity of wind power in China,the scale of the wind power cluster grid connection continues to increase,which poses great challenges to the safe and stable operation of the power system.Accurate wind power prediction can solve the uncertainty of wind power and increase the scale of wind power grid connection.Traditional wind power forecasting methods take into account simpler factors,their forecasting accurac y cannot be greatly improved,so there is an urgent need for new wind power forecasting theories and methods,which can improve wind power by studying the spatial correlation between wind farms.Forecast accuracy.To this end,this paper focuses on wind power prediction based on spatial correlation.The main research work is as follows:(1)Spatial correlation analysis of regional wind power.In this paper,the correlation coefficient is used to study the factors that affect the spatial correlation of wind power from both distance and time.The actual changes of wind power in wind farm in time and space are studied by the actual data of wind farm.Variational mode decomposition model can decompose wind power into different sub-models at two different scales of time and frequency,and improve the prediction accuracy by studying each sub-model.(2)Double-layer cluster analysis of wind farms based on GK-KFCM.In order to select a suitable reference wind farm,this paper proposes a two-layer clustering algorithm for wind farms based on spatial correlation and GK-KFCM.On the basis of selecting the main influencing factors of wind power,the influence of average wind farm availability on wind power is considered.Combined with the example data,Matlab software was used for simulation analysis,and the effective index was used to determine the number of clusters and the number of iterations.(3)Research on wind power prediction based on spatial correlation.In order to improve the accuracy of wind power prediction,this paper proposes the ICS-ELM power prediction model based on Chapters 2 and 3.The actual data of a wind power cluster in the northeast region were used for simulation experiments.Simulation examples verified the effectiveness of the method.The proposed model has improved to varying degrees compared to the prediction errors of traditional methods.