Research On Short-term Wind Speed Prediction Of Wind Farm Considering Chaotic Characteristic Of Atmospheric Motion

With the development of economic society and the increasing demand for energy,the development of wind power generation has been accelerated.The installed capacity and proportion of wind power increase rapidly.Considering the fluctuation of wind,the largescale wind power integration brings severe challenges to the safe and stable operation of power system.Short-term wind speed prediction is important for dealing with the uncertainty of wind power and promoting the consumption of wind power.Current research on short-term wind speed prediction is based on modeling and prediction to historical wind data,and rarely considers the physical characteristics of wind speed,which makes it difficult to further improve the prediction accuracy of short-term wind speed.To improve the accuracy of short-term wind speed prediction of wind farms,a decomposition method suitable for processing short-term wind data is studied based on the fractal characteristics of modes.Considering the influence of the chaotic characteristics of atmospheric motion on the actual wind speed,the chaotic characteristics of atmospheric motion is approximately represented,and the process of decomposition to wind speed is optimized.When the wind speed data are effectively decomposed,a prediction method with simple structure and more suitable for short-term wind speed prediction is studied.A short-term wind speed prediction method of wind farms considering the chaotic characteristics of atmospheric motion is proposed by integrating above methods.According to the analysis on the actual wind speed data collected from two wind farms,the effectiveness of the method proposed by this paper is verified by comparing with various methods.The main contents of the paper are as follows:(1)For the fluctuation and nonlinearity of wind speed,an adaptive EEMD decomposition method based on fractal characteristics is proposed.The influence mechanism of the amplitude of white noise and the number of ensemble on decomposition effect of EEMD is analyzed.Due to the different modes with different fractal dimensions,the wind speed can be accurately decomposed by adopting particle swarm optimization algorithm to find the parameters under minimum fractal dimension.In addition,the way of partial ensemble averaging is employed to improve the rapidity of the adaptive EEMD method by judging the separation state of abnormal signals from signals.The results show that the proposed method improves the dependence to parameters for EEMD,obtains more stable high-frequency components,and has higher degree of orthogonality between modes.(2)When appropriate parameters of white noises are determined,the decomposition effect of adaptive EEMD depends on the variation information in wind speed.Considering the influence of the chaotic characteristics of atmospheric motion on the actual wind,a wind speed decomposition optimization method considering the chaotic characteristics of atmospheric motion is proposed.The Lorenz-Stenflo equation is adopted to characterize the characteristics of atmospheric motion,and describes the chaotic information in wind speed.And the Lorenz-Stenflo equation is employed to optimize the wind speed decomposition process of adaptive EEMD,and reduces the adverse effects of atmospheric motion on wind speed.The effectiveness of the method on different wind speed signals is realized by judging the chaos degree of high frequency modes.The results show that the proposed method can reduce the adverse effect of atmospheric motion on wind speed by quantifying the influence of atmospheric motion,and obtain more regular modes compared with adaptive EEMD decomposition method,which is beneficial to the modeling of prediction methods.(3)To establish a wind speed forecasting model with simple structure and high precision,a continued fraction forecasting method with more general structure is proposed.Based on the contrast quotient theory,a more general continued fraction structure is deduced.The reliability of the continued fraction prediction is proved,and the structur e parameters of the continued fraction are estimated by the stepwise parameter estimation method.According to the fluctuation characteristics of wind speed,the prediction results are corrected by the fluctuation residual correction function.The results show that the proposed method can better capture the variation pattern of wind speed data and has higher prediction accuracy compared with five benchmark methods.(4)Combining the proposed wind speed decomposition method and wind speed prediction method,and considering the influence of the chaotic characteristics of atmospheric motion on short-term wind speed,a short-term wind speed prediction method considering the chaotic characteristics of atmospheric motion is proposed.The wind speed is decomposed into several modes by adaptive EEMD method optimized by considering the chaotic characteristics of atmospheric motion,and the predicted results of the original wind speed are obtained by superimposing the prediction results of each mode component by the continuous fraction method.To analyze the prediction performance of the proposed method,wind speed data with two time scales from two wind farms are analyzed.The results show that the proposed method combines the advantages of respective methods,and has obvious advantages in prediction accuracy.