含大规模风电场的电力系统运行优化方法研究

发布时间：2018-05-16 08:09

本文选题：风力发电 + 点估计　；参考：《华中科技大学》2014年博士论文

【摘要】：随着煤、石油、天然气这些化石能源的日益枯竭及其对环境污染的严重影响,作为可再生清洁能源中技术最为成熟的风力发电得到了大规模的发展。然而随着风电在电力系统中渗透率的不断增加,风电所具有的强波动性和不确定性将给电力系统调度运行带来越来越重大的影响,传统的电力系统调度运行方法可能不再适用,因此必须针对风力发电自身的特点,建立适用于含大规模风电电力系统优化调度运行的模型,并研究相应的求解和评估方法。本文首先回顾了电力系统优化运行相关的机组组合、经济调度和潮流计算三个关键问题的研究历史与现状,在此基础之上,针对风电对这三个问题产生的影响特性,重点研究了考虑风电爬坡事件的机组组合问题、计及机组调节能力的含风电电力系统的经济调度问题和基于风电数据样本的概率潮流问题。主要工作包括以下3个部分： (1)在含风电的电力系统机组组合问题中,针对风电功率可能会出现较大爬坡事件这一问题,建立了考虑风电爬坡事件约束的机组组合模型,并首先通过精确线性化技术将随机非线性混整模型转化成随机线性混整模型,然后考虑用风电预测值以及区间预测上下限来描述其出力,利用线性鲁棒优化理论将随机线性混整模型转换成确定性线性混整模型,最后可利用商业软件对其进行大规模求解。仿真结果表明了该方法的可行性和有效性,且对于一般的风电爬坡事件,爬坡事件约束是不起作用的,但对于风电波动速率较大的情况,考虑爬坡事件更能保证系统安全。 (2)在含风电的电力系统经济调度问题中,针对风电功率难以精确预测,在风电预测值的基础上,考虑风电预测误差上下限,建立了计及机组调节能力的含风电电力系统的鲁棒经济调度模型,其中风电预测不准所引起的误差可通过常规机组的调节能力得到线性最优补偿,应用线性鲁棒优化理论可将随机问题转化为确定性问题后利用商业软件进行求解,并保证系统安全,为含大规模风电的电力系统经济调度提供种新的有效方法。 (3)在含风电的电力系统概率潮流问题中,针对风电功率分布特性难以用常见的概率密度函数进行拟合,提出了一种基于风电样本数据的点估计方法,并在此基础上利用Cholesky分解技术处理输入变量的相关性,最后利用Gram-Charlier级数展开得到输出随机变量的累积分布,详细分析了不考虑与考虑样本数据相关性对仿真结果的影响。仿真结果表明该方法计算量小,精度高,且在实际工程应用问题中应考虑输入变量的相关性,而忽略其相关性可能会造成较大的误差。
[Abstract]:With the increasing depletion of fossil energy such as coal, oil and natural gas and its serious impact on environmental pollution, wind power generation, which is the most mature technology in renewable clean energy, has been developed on a large scale. However, with the increasing permeability of wind power in power system, the strong volatility and uncertainty of wind power will bring more important impact to the dispatching operation of power system, and the traditional dispatching operation method of power system may not be applicable. Therefore, according to the characteristics of wind power generation, a model suitable for optimal operation of wind power system with large-scale wind power should be established, and the corresponding solution and evaluation methods should be studied. This paper first reviews the research history and present situation of three key problems related to optimal operation of power system, such as unit combination, economic dispatch and power flow calculation. Based on this, the influence characteristics of wind power on these three problems are discussed. The problems of unit combination considering wind power climbing event, the economic dispatching problem of wind power system with unit regulation ability and the probabilistic power flow problem based on wind power data sample are studied in detail. The main work consists of the following three parts: 1) in the power system unit combination problem with wind power, in order to solve the problem that the wind power may occur a large slope climbing event, a unit combination model considering the wind power climbing event constraint is established. Firstly, the stochastic nonlinear blending model is transformed into the stochastic linear blending model by the exact linearization technique, and then the wind power prediction value and the upper and lower bounds of interval prediction are considered to describe the output force. The stochastic linear blending model is transformed into a deterministic linear blending model by using the linear robust optimization theory. Finally, it can be solved on a large scale by commercial software. The simulation results show that the method is feasible and effective, and that the constraint of climbing event is not effective for general wind power climbing event, but for the case of large wind power fluctuation rate, the system safety can be ensured by considering the climbing event. (2) in the economic dispatch problem of power system with wind power, considering the upper and lower limits of wind power prediction error, it is difficult to predict wind power accurately on the basis of wind power forecast value. A robust economic dispatching model of wind power system with unit regulation capacity is established, in which the error caused by the uncertainty of wind power prediction can be compensated linearly and optimally by the regulation ability of conventional units. By applying linear robust optimization theory, stochastic problems can be transformed into deterministic problems and solved by commercial software, and the system security can be guaranteed, which provides a new effective method for economic dispatch of power systems with large-scale wind power. In the probabilistic power flow problem of power system with wind power, a point estimation method based on wind power sample data is proposed to solve the problem that the distribution of wind power is difficult to fit with the common probability density function. On this basis, the Cholesky decomposition technique is used to deal with the correlation of input variables. Finally, the cumulative distribution of output random variables is obtained by using Gram-Charlier series expansion. The effect of not considering and considering the correlation of sample data on the simulation results is analyzed in detail. The simulation results show that the proposed method has the advantages of low computational complexity and high accuracy, and the correlation of input variables should be considered in practical engineering applications, while ignoring the correlation may result in large errors.
【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM614

【参考文献】