Research On Short Term Wind Power Prediction Based On EEMD-ANFIS

Due to the nonstationary and volatility of wind energy,grid-connected wind power on a great scale will bring a great impact on the power grid.Accurate prediction of wind power is not only conducive to the power grid scheduling,but also can improve the efficiency of wind power utilization and ensure the stable working of the power grid.This paper starts from the wind data used for wind power prediction,revising the imperfect part of wind data is helpful for reducing the prediction error from the source;then,not only is the algorithm considered,wind power prediction model is also established to realize short-term wind power prediction according to the data characteristics of wind speed and wind power,and main research contents are as follows:Firstly,a series of studies are found based on measured wind speed and wind power data in wind farm: wind speed and wind power are very random data;they exhibit nonlinear and non-stationary characteristics,meanwhile,the largest Lyapunov exponent of wind power indicates that the data has the characteristics of chaos.Secondly,as for missing and abnormal data problems in the measured wind data,the Least Squares Support Vector Machine(LSSVM)model which is optimized by the Comprehensive Learning Particle Swarm Optimization(CLPSO)algorithm is used,namely CLPSO-LSSVM model,and it is taken as revising model of the imperfect data,the correlation coefficient is taken as evaluation index,and it is compared with the revising results of the common models,the validity of this revising model is verified.Finally,the prediction model of EEMD-ANFIS is established according to the non-stationary and nonlinear characteristics of the wind speed and wind power data.Firstly,the ensemble empirical mode decomposition(EEMD)algorithm is used to decompose the data into a series of relatively stable components,then the Adaptive Neural Fuzzy Inference System(ANFIS)prediction model of each decomposed component is established,finally,the prediction results of each component are summed.Also,the prediction model is established based on EEMD-ANFIS by respectively combining with the effects of meteorological factors on the wind power,wind power of chaotic characteristics,wind speed and wind power conversion curves,the precision levels of three models are evaluated to determine the best prediction model.