高比例风电并网提高消纳能力的研究

发布时间：2018-01-12 09:27

本文关键词：高比例风电并网提高消纳能力的研究　出处：《沈阳农业大学》2017年博士论文　论文类型：学位论文

【摘要】：现阶段,雾霾等气候环境问题日益突显,常规化石能源储量有限等环境和能源问题已成为全世界普遍关注的热点课题。各个国家和地区都在倡导能源转型和能源结构改革,我国可再生能源发展"十三五"规划也明确提出逐步实现由以化石能源消耗为主的高碳能源模式向以可再生能源为能源消费主体的低碳能源模式过渡,促进能源转型。近几十年,全球都在提升可再生能源的发展与利用,世界及我国风力发电装机容量逐年攀升,从2010年至今我国风电总装机容量一直处于世界首位。但是,我国风电消纳却未能居于世界前列。选题立足高比例风电并网背景下,以提高风电消纳能力为研究目标,以不同时空尺度的风电运行特性分析为先导,凝练了风电功率预测研究与风电储能研究两个主要研究方向开展相关研究工作,即提高风功率预测精度从可靠性方面提高电力系统对风电消纳,储能配置是从调峰角度提高风电消纳。以大数据分析为研究的总体思路开展不同时空尺度的风电运行特性分析研究,归纳了风电场数据分析的层次结构,风电场数据分类框架结构及不同时空尺度风电运行特性分析架构体系。对风电场的月度和日频以及风力发电机的日频和十分钟运行数据进行了收集整理,开展了风电场月发电量季节性分析,风电场日发电量波动性、相关性以及日限电率分析。对风力发电机的日均风速、日发电量和日平均风速与日发电量相关性、日损耗电量与日平均温度相关性,以及风力发电机十分钟风速与功率数据进行了分析,并利用中位数区间估计方法绘制了实际运行风功率曲线。在保证系统电力电量平衡的前提下,通过分析研究确立了长时间尺度以发电量研究为主有助于风电消纳规划,短时间尺度以输出功率研究为主以确保风电的实时调度消纳。同时,通过对风电数据的季节性、相关性、波动性及分布特性及限电率等指标进行研究分析,进而明晰了论文从四个具体研究方向实现提高风电消纳能力研究,分别是风电场月发电量预测研究,风力发电机超短期输出功率预测研究,风电场/风电集群集中储能容量确定研究和基于就地储能的储发一体风力发电系统研究。立足预测波动研究,引入计量经济学和非参数统计学的分析和建模方法,开展中期和超短期风电功率预测方法的研究。利用季节变动时间序列模型、Holt-Winter季节性指数平滑模型、SARIMA季节时间序列模型以及基于熵值法的组合预测模型对风电场月发电量数据进行建模预测,通过对预测结果的评估分析验证了预测方法的适用性和合理性。另外,利用ARIMAX引入外生变量的时间序列模型进行风力发电机超短期风功率预测研究,并从确定性点预测扩展到不确定的概率区间预测,提高了预测的可信度,为风电实时调度消纳提供有效的参考。以平抑风电波动,提高风电消纳以及探索"源-网-荷-储"运营模式为研究目标,开展风电储能研究。对现行电力系统的储能方式进行了归纳分析,提出了按照大规模集中储能、小容量分散储能和即发即储就地储能三种储能布置方案及其结构。研究了消纳最大化和经济性分析相结合的确定风电场/风电集群的钒液流电池储能系统容量方法。另一方面,采用以平抑波动为目标来确定风力发电系统的就地储能布置方案,即从源头遏制风电出力的波动性,提出了储能前置和储能后置两种储发一体风力发电系统模型结构,分析了两种储能模型串并联结构的特点,并分别设计了储能前置和储能后置两种储发一体风力发电系统模型。综上所述,在高比例风电并网背景下,以风电运行特性为依据,以提高风电消纳能力目标,研究了风电场月发电量建模及预测,风力发电机超短期风功率建模及预测,基于钒液流电池的风电场/风电集群储能容量确定方法和基于就地储能的储发一体风力发电系统模型,为提高电力系统风电消纳探索了行之有效的技术方法。
[Abstract]:At present, the haze problem of climate and environment increasingly, conventional fossil energy reserves is limited and other environmental and energy issues have become a hot topic of common concern of the whole world. Various countries and regions are advocating the transformation of energy and energy restructuring, energy development planning "13th Five-Year" China Renewable also clearly gradually realize the high carbon energy mode from the consumption of fossil energy transition to low carbon energy to renewable energy as the main energy consumption, promote energy transformation. In recent decades, the global increase in development and utilization of renewable energy, the world and China's wind power installed capacity increased year by year, since 2010 China's total installed capacity of wind power has been in the first in the world. However, China's wind power consumption but not in the world. Based on the topic of high proportion of wind power under the background, in order to improve the absorptive capacity of wind power for research purpose Standard to wind power operation characteristics of different space-time scale analysis for the pilot, the refining of wind power prediction of two main research directions to carry out research work of wind power storage, improve the prediction accuracy of wind power from aspects of improving the reliability of wind power consumptive power system, energy storage configuration from the angle to improve the wind load power consumption. To analyze large data analysis of wind power operation characteristics of different spatial and temporal scales as the research framework, summarizes the structure analysis of wind field data, analysis system of wind farm data classification framework and different space-time scale wind power operation characteristics. On the frequency of wind farm and wind on monthly and frequency the generator and ten minutes of operation data were collected, carried out analysis of the season wind farm power generation, wind power generation, volatility, correlation and Analysis on power rate The daily average wind speed of the wind turbine. The generation of correlation, and daily average wind speed and generation, daily consumption and daily average temperature dependence, and wind power generators ten minute wind speed and power data are analyzed, and the median interval estimation method of drawing the wind power curve of the actual operation. Under the premise of ensuring the balance of power system next, through the analysis and research has established long time scale in power generation research mainly contributes to the wind power planning, the short time scale with output power of research to ensure the real-time scheduling of accommodating wind power. At the same time, through the season, the wind power data correlation, volatility and distribution characteristics and power rate etc. the index of research and analysis, and then clear the papers from the four specific orientation of research to improve the absorptive capacity of wind power, are the prediction of monthly power generation capacity of wind farm, wind power generation Prediction of machine of ultra short term power output of wind farm cluster / wind power centralized storage system to determine the capacity of research and the research of wind power generation system in energy storage. Based on the one based on the volatility forecasting research, analysis and modeling method of introducing econometrics and non parametric statistics, research on medium and ultra short term wind power prediction method the use of seasonal variation. Time series model, Holt-Winter seasonal exponential smoothing model, SARIMA model and seasonal time series forecasting model based on entropy method of wind farm power generation data modeling and forecasting, through the evaluation and Analysis on the prediction results validate the prediction method of applicability and rationality. In addition, the time series model by introducing ARIMAX exogenous wind turbine ultra short term wind power prediction research, and from the deterministic point prediction is extended to probabilistic interval prediction uncertainty Test, improves the reliability of prediction, provide effective reference for wind power consumption. In order to stabilize the real-time scheduling of wind power fluctuation, improve the wind power and the exploration of the source network load - reservoir operation mode as the research object, carry out the storage of wind power research. The current power system for energy storage the inductive analysis, put forward according to the large-scale centralized storage capacity, small dispersed storage and send local storage storage three storage layout and structure. Study on Determination of vanadium in wind farm cluster / wind power flow battery energy storage system capacity for consumptive maximization and economic analysis of the combination. On the other hand, used to stabilize fluctuations as the goal to determine the local storage for wind energy generation system layout, from the source to curb the fluctuation of wind power output, puts forward energy storage and storage two pre post storage one model of wind power generation system, analyzed two kinds of The storage model on the parallel architecture, and designed respectively storage and storage two pre post storage one model of wind power generation system. To sum up, in a high proportion of wind power in the background, the operational characteristics of wind power as the basis, to improve the absorptive capacity of wind power generation, wind farm modeling and June study on prediction, ultra short term wind power generator wind power modeling and prediction of wind / vanadium flow battery for wind power cluster storage capacity determination method and based on the local energy storage one wind power generation system based on the model, in order to improve the power system of wind power consumptive explores an effective method.

【学位授予单位】：沈阳农业大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TM614;TM73

【参考文献】