| Due to the fossil energy crisis and environmental pollution problems,the research and development of renewable energy technology such as wind power has become the developmental trend of the electrical industry.As a relatively mature renewable energy,wind power is clean,green,and abundant However,due to its inherent uncertainty,the integration of wind power has brought many problems to the power grid.Therefore,it is of great significance to investigate the uncertainty characteristics of wind power and their countermeasures regarding the wind power consumption and power grid security and stability.If the uncertainty of wind farm is transferred to the power grid,the impact of uncertainty will be expanded from the generation side to the whole power system,and the complexity of the planning and regulation of the whole power system will be increased.If the wind farm itself has a certain degree of self-discipline,the dispatchability of wind power will be enhanced while the complexity of the whole system will be reduced.Therefore,it is reasonable to enhance the autonomy of wind farms and enable them certain self-discipline.An energy storage system with appropriate capacity is installed at the outlet of the wind farm so that the wind farm cooperates with the energy storage system.Through this cooperation,the wind farm obtains certain self-discipline ability,which leads to the enhanced friendliness and dispatchability of the wind farm,and the role transformation from passive acceptance to active participation in absorption.Therefore,it is of great practical significance to study the optimal energy storage capacity allocation method of autonomous wind farm.In addition,the uncertainty of wind power is actually an irregular distribution.Any researches on wind power forecasting and energy storage capacity optimal allocation based on regular distribution have limitations,which do not have the robustness of the research model and the universality of research methods.Therefore,in view of the irregular distribution characteristics of wind power uncertainty,it is of great theoretical significance to find a robust research model and general research method.The objective of this paper is the autonomous wind farm with energy storage.We first analyzed the uncertainty characteristics of wind power and the wind power short-term interval prediction under irregular distribution.Then,based on the prediction of wind power interval,the optimal allocation of energy storage capacity of autonomous wind farm is investigated from the following three aspects:(1)From the perspective of wind farm self-discipline interval optimization,the optimal allocation of energy storage capacity is studied;(2)From the perspective of risk analysis,the optimal allocation of energy storage capacity is studied;(3)In frequency domain,analyzing the wind power frequency characteristics,from the perspective of smoothing the wind power fluctuation.,the optimal allocation of energy storage capacity is studied.Eventually,the research system of autonomous wind farm from wind power uncertainty analysis to wind power interval prediction and to energy storage capacity allocation is established.The model robustness and the method universality are considered.In actual wind farm,the optimal allocation of energy storage capacity can be determined from the above-mentioned angles or some of them according to the actual situation of units,regions and markets.The main research work and innovation of this dissertation are as follows:(1)According to the irregular distribution characteristics of wind power,a wind power interval prediction method based on Parzen window and confidence interval optimization is proposed.Firstly,the Parzen window estimation method is used to establish the uncertainty model of wind power.The Parzen window estimation method is suitable for irregular distribution of arbitrary shape.Secondly,the solving of the prediction interval of irregular distribution.This dissertation takes the minimum prediction interval width as the objective function and satisfies a certain constraint conditions,establishes an optimization model and applies the trust region optimization method to solve the optimal prediction interval.Finally,regarding the evaluation of interval prediction quality,interval width and interval prediction accuracy are two contradictory and mutually restricting indicators.This dissertation adopts the concept of F value from the field of information retrieval to comprehensively measure the two indicators,so that to comprehensively and objectively evaluate the prediction quality.Case study results show that this method not only can obtain better quality of interval prediction,but also has model robustness and method universality.It lays a solid foundation for the following chapters on the optimal allocation of energy storage capacity of autonomous wind farm.(2)From the perspective of wind farm self-discipline interval optimization,an optimal allocation method of energy storage capacity is proposed based on tracking the optimal self-discipline interval and effectively tracking the planned output.1)The wind farm tracks the optimal self-discipline interval.Firstly,the concept of self-regulation interval of wind farm is proposed.The actual output of wind farm can be limited within a certain power range by configuring energy storage in wind farm,so as to reduce the impact of wind power uncertainty on power grid and enhance the schedulability and friendliness of wind power.Then,according to the irregular distribution characteristics of wind power uncertainty,try to find the optimal self-discipline interval and the corresponding optimal allocation of energy storage capacity.2)The wind farm effectively tracks the planned output.If the wind farm can strictly track the planned output(at this time,it can be regarded as the case that the width of self-discipline interval is zero),the wind farm will have stronger self-discipline.But at this time,the wind farm needs more energy storage capacity,so the economy must be considered.Therefore,the wind farm can consider tracking the planned output under a certain tracking degree.For wind curtailment and wind shortage beyond the tracking degree,the wind farm should pay the corresponding wind curtailment penalty and wind shortage penalty to the grid to purchase the spinning reserve of the system to maintain its planned output.Taking the maximum profit of wind farm as the objective function,the optimal allocation of energy storage capacity under certain tracking degree is solved for irregular distribution.On this basis,the relationship between the optimal result and the influencing factors such as energy storage cost,electricity price,wind curtailment penalty,wind shortage penalty and tracking degree are analyzed in detail.Simulation results show that the method is not only suitable for regular distribution,but also suitable for irregular distribution.It has robustness in research model and universality in research method.It can not only obtain the optimal allocation of energy storage capacity based on the optimal self-discipline interval of wind farm,but also obtain the optimal allocation of energy storage capacity under certain planned output tracking degree and the optimal tracking degree under different cost conditions and penalty conditions.The proposed method effectively improves the self-discipline level of wind farms,and provides a strong decision support for the energy storage capacity planning of autonomous wind farms,and has important reference value.(3)From the perspective of risk analysis,this paper investigates the optimal allocation of energy storage capacity of autonomous wind farm,and proposed the optimal allocation method of wind farm energy storage capacity based on improved Sharpe Ratio.The uncertainty of wind power will bring risks to the system.The autonomous wind farm should has a certain risk aversion ability.The wind farm's extra profit of per unit risk can be evaluated by using the idea of Sharpe Ratio in the financial field,organically combining both the risk and the profit.Firstly,the traditional Sharpe Ratio is improved.The Conditional Value at Risk(CVaR)is used to replace the risk measurement in the traditional Sharpe Ratio to measure the risk caused by wind power uncertainty.This is not only applicable to the irregular distribution characteristics of wind power uncertainty,but also takes into account the tail risk and is forward-looking.Secondly,considering that the wind power uncertainty will bring bilateral risks to the system(i.e.wind power curtailment risk and shortage risk),the traditional unilateral CVaR is improved and the calculation of bilateral risk under irregular distribution is extensively studied.Finally,by comprehensively measuring the unit risk and extra profit,and taking the maximum Sharp Ratio as the objective function,the optimal energy storage capacity allocation of wind farm will be solved.Case study results show that this method not only combines risk and profit,but also is suitable for arbitrary distribution.It has robustness in research model and universality in research method.It provides a new idea for risk assessment and optimal allocation of energy storage capacity of autonomous wind farm(4)In frequency domain,analyzing the wind power frequency characteristics,from the perspective of smoothing the wind power fluctuation.,an optimal allocation method of hybrid energy storage capacity in autonomous wind farm is proposed,which considers the fluctuation rate margin of wind power low frequency component.In the frequency domain,the wind power is decomposed into low-frequency component,sub-high-frequency component and high-frequency component by wavelet packet transformation.The low-frequency component satisfied the grid connected fluctuation standard is taken as the grid connected power,while the sub-high-frequency component and high-frequency component are mitigated by battery and super capacitor respectively.Considering that there is a certain fluctuation margin of low-frequency components in grid connection,it is only necessary to partially mitigate the sub-high-frequency components and high-frequency components.Under the guidance of this idea,taking the minimum energy storage cost of wind farm as the objective function and the fluctuation as the constraint condition,a mathematical model is established to obtain the optimal capacity allocation of battery and super capacitor and the optimal confidence degree.Furthermore,by analyzing the entropy of each typical daily energy storage allocation result and the information content and uncertainty,and the entropy weight of each typical daily result is determined.Based on the results of each typical daily energy storage allocation result and its entropy weight,the final optimal allocation result of hybrid energy storage is obtained.Case study results show that the proposed method can significantly reduce the energy storage cost under the premise that the wind power fluctuation meets the grid connected standard,and the results have smaller power error and capacity error,which effectively improves the accuracy and rationality of energy storage capacity allocation. |
PDF Full Text Request