Uncertainty Of Wind Power Forecasting And Robust Optimal Scheduling Of Power Systems With Wind Power

As a clean energy source,wind energy has been widely used in countries around the world.However,the volatility of wind resources itself determines the uncertainty of wind energy utilization.Whether it is land wind power or offshore wind power,the wind power forecasting uncertainty will be faced when using wind energy.Therefore,how to improve the wind power power prediction accuracy is one of the problems that wind energy utilization needs to solve.In addition,the large-scale wind power integration into the power grid will also have a great impact on the power system scheduling.How to improve the wind energy utilization rate and reduce the abandonment Wind volume,reducing the power generation cost of power systems,improving power reliability and power quality are another major problem to be solved by wind energy utilization.Therefore,this paper proposes a wind power prediction method based on support vector machine,and studies the uncertainty of wind power prediction,and constructs an uncertain set suitable for power system robust optimization scheduling and the robustness of power system with wind farm.Optimize the scheduling model.The main research contents are:1.Wind power prediction based on support vector machine(SVM).The wind speed characteristics and wind direction characteristics of wind farms and the support vector machine are studied.The actual wind farm data is used to select the appropriate samples for sample selection.The support vector machine prediction model is used for prediction,and the predicted results are passed.The results obtained by RBF and wavelet analysis are compared and analyzed.Finally,the example analysis shows that the support vector machine prediction method has higher prediction accuracy.2.This paper analyzes the uncertain factors affecting the prediction results,introduces confidence intervals to describe the wind power uncertainty,and compares the confidence intervals of the three units at different confidence levels to reflect the fluctuation range of the prediction results;3.By constructing the uncertainty analysis of the prediction results to construct an uncertain set that can be used for power system scheduling,the uncertain set of prediction errors constructed is an elliptical set,which can greatly reduce the degree of conservatism while ensuring the accuracy of the set.Finally,the convergence is verified by simulation analysis.The results show that the constructed elliptic set still has good convergence under the condition of reducing the conservativeness.4.According to the possible scenarios of wind power in the uncertainty set,rationally plan the output of the automatic power generation control(AGC)unit,establish a robust optimization economic dispatch model with the goal of minimizing power generation cost and maximizing the use of wind power,and use the power plant in a certain area.The data is analyzed as a case study.The results show that the constructed electroplating model has good robustness,reliability and economy.