Quantitative Method Research On Power Clustering Effect Of Wind Power Cluster

The shortage of traditional energy and the increasingly severe energy and environmental crisis make it an inevitable choice to develop renewable energy to achieve sustainable development.As the most commercialized renewable energy,wind power generation has been developing continuously and rapidly,and large-scale wind farm cluster transmission has become a trend.At present,our country has wind energy resources are rich regions such as Jiuquan,Xinjiang Hami,ten 10 GW grade wind power bases have been built,with the increasing of wind power grid capacity,wind power show that the clustering effect,cluster analysis and grasp the characteristics of the wind clustering effect for large scale wind power grid operation is of important guiding significance.This paper takes wind power clustering effect as the entry point to study the trend of wind power clustering effect as follows.Firstly,the fluctuation characteristics and clustering effect of large-scale wind power are analyzed.Based on the statistical analysis index of wind power fluctuation characteristics,the fluctuation characteristics of output power of a single wind turbine,a single wind farm and the whole wind farm group are analyzed.By comparing the characteristics of wind power output of different scales,the wind power clustering effect is derived and the trend of wind power clustering is analyzed from the perspective of statistical analysis and time-frequency analysis.Secondly,the quantitative method of wind power clustering tendency based on statistical analysis is studied.The method of state division of wind power continuous output curve was given,and a combined prediction model of wind power continuous output curve state based on improved shapley value was established.In order to test the accuracy of the prediction model,the prediction accuracy index evaluation system was constructed.An example is given to compare and analyze the traditional shapley value combination prediction model,three single prediction models and the method in this paper,and the influence of the number of state division on the prediction accuracy of the model is analyzed,which verifies the correctness of the model and the validity of the proposed method for the description of wind power clustering trend based on statistical analysis.Finally,the quantitative method of wind power convergence tendency based on time-frequency analysis is studied.In this paper,the idea of applying db6 discrete wavelet to the analysis of convergence effect is introduced.According to the typical peak regulation and frequency regulation time scale of the system,the wind power frequency band division method for power grid operation is presented.In order to quantify the time-frequency characteristics ofwind power clustering tendency in each frequency band,the time-frequency domain analysis index system of clustering effect is constructed.The example is divided into two parts.In the first part,the trend of wind turbine cluster clustering effect is analyzed quantitatively,and the influence of random sampling on the conclusion is discussed.The second part studies the trend of low-frequency wind power output in a larger clustering scale.The analysis shows that with the increase of wind power clustering scale,the influence of wind power in each frequency band on the power grid decreases exponentially.The research work in this paper can provide a basis for large-scale wind power grid operation and planning.