Method For Predicting Monthly Wind Power Generation

New energy power generation,especially wind power generation,has developed rapidly in China in recent years.At present,China has the largest installed capacity and newly added wind power capacity in the world.The output power of wind power has a strong uncertainty,so compared with the traditional fossil energy generation,the controllability is low,and it can not produce the required power according to the transaction plan like the traditional power.China's monthly electricity trading plan formulation process is to give priority to the allocation of wind energy,and the rest to thermal power units.With the increasing proportion of wind power generation capacity,if the original approximate estimation method is still used to develop the monthly wind power trading plan,the inaccurate monthly wind power trading plan may cause the corresponding monthly thermal power trading plan difficult to implement,seriously affecting the safe and stable operation of the grid.In this context,how to effectively and accurately predict the monthly wind power generation becomes the key to ensure the safety and reliability of the power system.At present,there are two difficulties in the monthly forecast of wind power generation at home and abroad,which are insufficient data and inaccurate weather forecast information.Generally,the construction time of wind farms is short,and the monthly forecast must take the monthly power generation as a whole as a data volume,resulting in the lack of data needed for the forecast;the wind power generation has a direct relationship with weather information,but due to the long time scale of monthly forecast,the accuracy of weather forecast information will gradually reduce,which can not meet the forecast requirements.Firstly,Pearson and multiple linear regression analysis are used to select the maximum wind speed,the minimum wind speed and the temperature of a single day as the characteristics of the prediction.Aiming at the problem of large error caused by the lack of historical data,according to the strong seasonal characteristics and the characteristics of short-term smooth change of wind power,an extension technology of monthly wind power generation data set is proposed,which is mainly divided into historical data to supplement In order to provide sufficient data support for the later prediction,two parts are added to the monthly forecast data.Aiming at the problem that the monthly forecast of wind power generation faces the uncertainty of weather information,an entropy weight combination comprehensive forecast method based on the three prediction algorithms of unit matching,data expansion and time series is proposed by using the expanded historical data,and a data correction method considering maintenance is proposed To improve the prediction accuracy.The simulation data analysis shows that the proposed data supplement method is feasible and effective;the comprehensive prediction method is not only accurate,but also can avoid the influence of a single month prediction error of a certain prediction method on the overall prediction results included in the maintenance data,which can effectively improve the accuracy and stability of the whole prediction.