Research On Short-term Wind Power Forecasting And Its Uncertainty Based On Relevance Vector Machine

In recent years,due to serious environmental pollution,the problem of the earth's ecological environment has become increasingly prominent.The development of clean energy has attracted widespread attention from the society.The development of new energy power generation based on wind power generation has been rapidly developed globally.However,the randomness and volatility of wind energy make wind power forecasting uncertain,and accurate and reasonable forecasting can make the power system reliable,continuous and stable.The paper proposes a short-term wind power forecasting method based on the relevance vector machine to study the short-term forecast and uncertainty of wind power.The main research content includes:1.Introduce four kinds of statistical models(time series method,neural network method,support vector machine,correlation vector machine)for short-term forecasting of wind power respectively.Through the comparison of four predictive models,it can be known that the correlation vector machine is for small sample data.Have better predictability.And analyzed the basic characteristics of wind speed and screened the wind speed data.2.Short-term wind power forecasting based on Relevance Vector Machine(RVM).Because of the volatility and randomness of wind speed,in the process of analysis and modeling,the actual wind farm data needs to take into account changes in the weather,possible errors in the numerical weather forecast,and other issues,and correct the data.Then the correlation vector machine is used to predict the wind power,and better prediction results are obtained.3.Analyze and study the influencing factors of forecasting uncertainty of wind power,divide the existing form and evaluation index of forecast error,and analyze the influence factors of three kinds of uncertainties(including data accuracy,unit failure,accuracy of prediction model).).In the analysis of the examples,the input data accuracy and accuracy of the prediction model are mainly considered in the influence of uncertainty.4.Wind power forecast and uncertainty analysis.Considering the uncertainty of wind power is the theoretical and model basis for modeling and analysis.The historical NWP and historical wind farm power data are used as training samples,and then combined with future numerical weather forecasts to predict the power generated by the wind farm.Through the analysis of the uncertainty,research methods to improve the accuracy and high efficiency of wind power uncertainties are of great significa nce for relieving the phenomenon of “wind rejection” of wind farms.