Theoretical Studies On Wind Power System Forecasting And Decision-making Based On Characteristic Mining Of Uncertainty

It is an important strategic move that replaces the fossil fuels with clean energy and achieves low-carbon and green development for primarily debasing the crisis of energy and resources,and solving the pollution of the environment what contemporary mankind facing.Recently,for the low carbon energy strategy in our country,wind power generation plays a leading role for coping with the energy and environmental challenges,which is one of the most mature technology in the renewable energy technologies.However,the wind generation shows the significant volatility and intermittency due to the effects of natural meteorological factors,makes the power grid regulation face rigorous challenges with wind power integration and directly affects the reliable and efficient utilization and the development of the future of wind power.The most important cause is that it is difficult to accurately grasp the uncertainty characteristics of wind power and sharply improve its forecasting accuracy,and results in the regulation mechanism difficult working effectively.Under the above-mentioned background,based on characteristic mining of wind power uncertainty and clearing wind variation rule for guidance,this dissertation gradually focuses on the studies of data analysis,forecasting and decision-making.First,mining the wind power characteristics is supported to improve the forecasting techniques and model the uncertainty,and based on that,to build a wind-storage system.The research system of fully addressing the uncertainty of wind power has been formed to cover the data analysis,building the model and decision-making.The key of this dissertation is to study the mechanism of reducing the uncertainty of wind power,to alleviate the pressure of wind power uncertainty on the power grid operation and finally to realize to achieve the large-scale wind power integrating into the grid.The main works and achievements of the dissertation are as follows.(1)Wind power exhibits extremely complex nonlinear dynamic characteristics,which are affected by many factors,such as wind speed,wind direction,air temperature,air humidity,illumination and earth surface roughness.A method for the scale division of wind power based on the Hilbert-Huang transform(HHT)and Hurst analysis is proposed in this paper,which allows the various multi-scale dynamic behavior characteristics of wind power to be investigated.First,the time-frequency characteristics of wind power are analyzed using the HHT,and then Hurst analysis is applied to analyze the stochastic/persistent characteristics of the different time-frequency components.Second,based on their fractal structures,the components are superposed and reconstructed into three series,which are defined as the Micro-,Meso-and Macro-scale subsequences.And the indices related to statistical and behavioral characteristics of the subsequences are calculated and used to analyze their nonlinear dynamic behavior.Finally,the very short-term wind power combined forecasting model based on the multi?scale analysis and end effect is proposed to demonstrate the effectiveness of the above-mentioned analysis method.(2)In this paper,combined with the relevant conclusions of the characteristics analysis,several novel short-term wind power combined forecasting models based on error forecast correction are proposed in the one-step ahead,continuous and discontinuous multi-step ahead forecasting modes.First,the correlation relationships of forecast errors of the autoregressive model,the persistence method and the support vector machine model in various forecasting modes have been investigated to determine whether the error forecast models can be established by regression learning algorithms.Second,according to the results of the correlation analysis,the range of input variables is defined and an efficient strategy for selecting the input variables for the error forecast models is proposed.Finally,several combined forecasting models are proposed,in which the error forecast models are based on support vector machine/extreme learning machine,and correct the short-term wind power forecast values.The data collected from a wind farm are selected as a case study to demonstrate the effectiveness of the proposed combined models.(3)It is very helpful for improving the safety and economy of the power system operation to accurately describe the statistical characteristics of wind power prediction error.On the basis of obtaining the day-ahead hourly multi-period forecast errors of wind power using the proposed short-term combined forecasting method,this paper proposes a method to model the wind power uncertainty of multi-period forecast error.The kernel density estimation method is used to describe the probability distribution of prediction error and the Pearson correlation coefficient and scatter diagram are utilized to analyze their correlation relationship.Then this paper proposes the modeling method of the joint probability distribution function for the prediction error using Copula theory,in which the marginal distribution and correlation relationship of the prediction error are both considered.Moreover,the marginal distribution functions of the prediction error of the wind power are without prior assumptions during the modeling process.The proposed model provides a theoretical basis for using the historical forecasting data comprehensively and accurately to model the uncertainty of wind power and decision-making of power system operation with uncertain information of wind power.(4)The fluctuation amplitude and speed of wind power are the key factors affecting the operation and decision of wind power system.For the multi-period optimization decision-making problem of charge and discharge power and capacity sizing of energy storage system that is used to compensate prediction error and track the plan output,on the basis of building the joint probability distribution function of the day-ahead hourly wind power prediction error,this paper proposes an optimal sizing method for the capacity and power of energy storage system based on the multi-scenario technique.This study realizes the optimal decision-making of sizing of energy storage system and the marginal distribution and the temporal correlation of the prediction error are also accurately described in this method.The test example demonstrates the effectiveness of the proposed model and the results show that the temporal correlation of prediction error has a significant effect on the capacity and power of the energy storage optimal sizing and if the temporal correlation is ignored,the capacity and power of the energy storage system would be seriously misestimated.