Research On Ultra-Short-Term Wind Power Interval Forecasting Method Based On Combined Model

In view of the increasing demand for electricity and the lack of traditional petrochemical resources,the installed capacity and power generation of clean resources,mainly wind power,are gradually increasing in recent years.Affected by the randomness and volatility of the wind itself,the electric energy generated by the wind turbine also has the same nature,showing strong uncertainty,especially when a lot of electric power from wind farm connecting to the power net.The imbalance of supply and demand in the power net greatly increases the complexity and pressure of peak regulation and frequency regulation of the power system,and brings difficulties to economic dispatch.The uncertainty modeling of wind power error to fit the its distribution provides a decision-making basis for the dispatch strategy,and the economic dispatch strategy arrangement of thermal power units considering the uncertainty of wind power forecast and wind power forecast comprehensively can ensure the power Under the premise of system safety and stability,the economic cost can be reduced to the greatest extent,and the problems caused by large-scale wind power grid connection can be effectively solved.Based on this,this paper does the following work:(1)Aiming at the problem that wind power has randomness,volatility and strong nonlinearity,which leads to poor forecasting effect of wind power,a wind power prediction method combining Complete Empirical Mode Decomposition-Variational Mode Decomposition(CEEMD-VMD)and Improved Grey Wolf Algorithm(IGWO)optimized least squares support vector regression(LSSVR)is proposed.The CEEMD-VMD quadratic decomposition overcomes the problem that the traditional CEEMD decomposition sub-components are too complex to make the model difficult to train.The improved gray wolf algorithm improves the full-dimensional optimization ability of the gray wolf algorithm,and further improves the prediction accuracy of the least squares support vector regression.For the prediction ability of vector regression,the model first performs CEEMD decomposition on the historical wind power sequence,quantifies its complexity by sample entropy,performs VMD secondary decomposition for the sub-components with excessive complexity,and finally applies IGWO-LSSVR to all sub-components respectively.The model predicts and superimposes the overall wind power forecast value.(2)A method for modeling the uncertainty of wind power based on the segmentation of the k-means clustering(k-means)-Gaussian mixture model(GMM)method is proposed..Firstly,clustering analysis is carried out using the k-means clustering method for the wind power prediction sequence and prediction error sequence;secondly,the Gaussian mixture model is used in different predicted value intervals to determine the distribution of the upwind power in different prediction intervals;finally,the Gaussian mixture model results On the basis of,the wind power interval prediction results are obtained by inverting the probability density function.(3)An economic dispatch model for thermal power units considering the uncertainty of wind power forecasting is studied.A dispatch model that considers both the operating cost of thermal power units and the uncertainty cost of wind power forecasting is constructed,and the nondeterminacy distribution method is constructed to convert the random variable model with nondeterminacy into a deterministic variable model.In terms of model solving,based on the sparrow algorithm with outstanding high-dimensional optimization ability,a sparrow algorithm integrated with the beetle algorithm is proposed to solve the problem of single population in the later stage of the sparrow algorithm to solve the scheduling model.The validity of the proposed algorithm and model is verified in the case analysis.(4)On the basis of the first three parts,considering the impact of thermal power plant group peak shaving technology on economic dispatch,explore the relationship between dispatch economic cost and peak shaving depth under different peak shaving depths,as well as different depths of peak shaving The relationship between the economic cost under the compensation strategy and the compensation of the deep peak shaving policy.Give rationalization suggestions.