Research On Ultra-Short-Term Wind Power And Short-Term Wind Power Prediction Based On Combined Model

As one of the new energy generation technologies,wind power generation has the advantages of clean,low-carbon,renewable.However,due to the characteristics of high randomness,strong uncontrollability and weak dispatchability of wind power.When wind power is incorporated into the power grid,the working efficiency of the power grid will be reduced,and It will also directly affect the safe and stable operation of the electricity grid.Therefore,effective prediction of wind power in wind farms is an important measure to better utilize wind energy resources,improve power quality after wind power integration,and increase the competitiveness of wind power in the electricity market.In this paper,the non-stationarity of wind power is considered,and the factors affecting wind power are analyzed in depth.Based on the time series nature of wind power and the effect of influencing factors,ultra-short-term and short-term prediction models of wind power are established.The main findings are as follows:(1)Three models for ultra-short-term prediction of wind power are established by time series analysis and signal decomposition.The results show that single ARIMA model has poor prediction effect.While empirical mode decomposition(EMD)algorithm can effectively improve the prediction accuracy,but it also causes the mode aliasing phenomenon.After adding white noise to establish EEMD-ARIMA model,the mode aliasing phenomenon is effectively optimized and the prediction accuracy is further improved.It is the most recommended ultrashort-term forecast model in this paper.(2)By analyzing the influencing factors of wind power,it can be seen that wind speed,as the main influencing factor of wind power,directly affects the output of wind power.Effective prediction of wind speed plays a positive role in improving the accuracy of wind power prediction.At the same time,in order to compare the impact of different prediction steps on the prediction accuracy of the model,further improve the prediction effect of the model,and enhance the applicability of the model under different scenarios and requirements.Based on the EEMD-ARIMA model,short-term multi-step prediction and short-term single-step prediction of wind speed and wind power are carried out.The results show that the multi-step prediction can not modify the data,so the prediction effect of the previous ten steps is relatively high.However,as the number of steps increases,the prediction results gradually deviate from the true value,or even the opposite trend,resulting in the larger prediction error.Single step prediction will constantly modify the training period data to ensure the accuracy of the prediction points.(3)In order to study the short-term wind power prediction more comprehensive,based on the characteristics of wind power series,this paper uses the Least Squares Support Vector Machine method to make short-term wind power prediction with wind speed as the influencing factor.At the same time,the PSO algorithm is used to optimize the regularization parameters and kernel parameters of the model.That the PSO-LSSVM model has a higher prediction accuracy and better prediction effect than the LSSVM model without parameter optimisation.M-PSO-LSSVM model and M-LSSVM model based on the results of short-term multi-step(M)and single-step(S)of wind speed have large errors in predicting wind power,and the accuracy is lower than S-PSO-LSSVM model and S-LSSVM model.(4)By analyzing and comparing the simulation results of EEMD-ARIMA,M-LSSVM,M-PSO-LSSVM,S-LSSVM,S-PSO-LSSVM,the prediction accuracy increases with the model.However,the prediction accuracy of M-LSSVM and M-PSO-LSSVM may be lower than that of EEMD-ARIMA.This is because the wind speed has a great influence on the prediction results of PSO-LSSVM model and LSSVM model.When the prediction accuracy of wind speed is low,EEMD-ARIMA model is recommended to predict short-term wind power;When the prediction accuracy of wind speed is high,the PSO-LSSVM model is recommended.