Research On Ultra-short-term Wind Power Prediction Based On CEEMDAN-PSR Combined With ISSA-LSSVM

To achieve the goal of "carbon peaking and carbon neutrality",my country has increased its research on new energy power generation.As an important part of new energy power generation,wind power generation has strong fluctuation,instability and randomness in its output power.Therefore,the grid connection of a large number of wind power will seriously affect the stable operation of the power system,and the power quality will also be affected.decline.In response to the above problems,this paper studies how to improve the ultra-short-term prediction accuracy of wind power generation,aiming to improve the stability and safety of power grid operation,ensure the power quality of the power grid,and help the power grid dispatching department to accurately formulate the wind power grid connection plan.The main work is as follows:First,in order to accurately estimate the variation trend of wind power output power,a Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)combined with Phase Space Reconstruction(Phase Space Reconstruction,PSR)wind power time series data processing method.First,the CEEMDAN method is used to decompose the wind power time series data to obtain a series of Intrinsic Mode Functions(IMF)that can characterize the characteristics of the original data.the preprocessed data.Then,aiming at the problem that the random initialization of individuals in the Sparrow Search Algorithm(SSA)may be too concentrated in the local area,the generality of the early search and the locality of the later search are unbalanced,and the problem of easy precociousness is proposed.The Tent map realizes the global random initialization of the algorithm population individuals,uses the compression factor to improve the way to guide the population update,and uses the Cauchy variation to perturb the improved sparrow search algorithm(Improve Sparrow Search Algorithm,ISSA),and the it is mathematically verified.The mathematical theory of Least Squares Support Vector Machines(LSSVM)is introduced in detail.At the same time,for the problem that the prediction accuracy of the LSSVM algorithm is greatly affected by the penalty factor and kernel function parameters,an ISSA is proposed.-LSSVM prediction model,using the ISSA algorithm to optimize the selection of the penalty factor and kernel function parameters of the least squares support vector machine,and effectively verify its prediction performance through experiments.Finally,the CEEMDAN combined PSR algorithm is used to preprocess the wind power data of 26 days from 0:00 on December 6th to 24:00 on December 31 st,2020 with the total power of 24 2MW wind turbines in a wind farm in Zhangbei,so as to obtain the phase space reconstruction.Then,the ISSA-LSSVM algorithm prediction model is established for each subsequence,and finally the final wind power prediction result is obtained by superimposing the prediction results of each reconstructed subsequence.Compared with other prediction models,the ultra-short-term prediction model of wind power based on the CEEMDAN-PSR combined ISSA-LSSVM algorithm proposed in this thesis has a very good prediction effect,which can provide a reference for specific engineering practice.The thesis has 40 figures,15 tables and 60 references.