基于潮流转移和追踪的含风电电力系统运行风险评估

为实现含风电电力系统运行风险快速评估,提出了一种基于潮流转移和追踪的风险快速评估方法。首先,考虑风电出力波动与时间和风速的相关性,基于Markov理论建立了风速相依的风电出力波动模型;其次,为提高系统状态分析效率,通过推导节点转移分布因子和适用于多支路开断的支路开断分布因子来实现快速潮流计算,避免了潮流迭代计算;然后,建立了基于潮流追踪的负荷削减模型,采用潮流追踪理论筛选出最有效的控制节点集合,将全系统范围内寻优转化为局部范围内寻优,在改进潮流计算算法和负荷削减模型2个方面实现了运行风险快速评估;最后,通过IEEE-RTS79系统的仿真分析,验证了所提模型的准确性和有效性,进一步计算切负荷风险指标和线路越限风险指标以分析不同风速条件对系统运行风险评估的影响,为含风电电力系统运行风险评估提供参考。     

基于风速时间周期特征的风电并网系统风险评估方法

目前的风电并网系统风险评估方法多采用风速的概率分布模型,评估的是系统全年的风险指标,不能反映风速和系统风险的时变特征。提出了风速的时间周期特征,并将其描述为风速长期、平缓的月变化趋势和短期、快速的日波动特征两部分的叠加。用时间周期拟合函数表示风速的月变化趋势,用服从特定概率分布的随机变量表示风速的日波动特征,通过对多年风速样本进行曲线拟合来建立风速的时间周期特征模型。根据该模型模拟得到的时变风速建立风电场出力模型,采用蒙特卡洛模拟方法计算风电并网系统中长期风险指标,反映了系统风险的时变特征。以IEEE-RTS79系统及某风电场实际风速为例,验证了所提方法的有效性。评估结果可为电力系统规划、中长期调度和月发电计划制定等提供重要参考。     

基于互Box-Cox变换和Markov链风速云模型的发电系统充裕度评估

为准确计及风速随机性和自相关性对风电并网系统充裕度的影响,建立基于互Box-Cox变换和Markov链的风速云模型,并将该模型与时序Monte Carlo模拟法结合,提出计及风速随机性和自相关性的风电并网系统充裕度评估方法。仿真结果表明,所提模型产生的仿真风速样本与实测风速样本具备相似的概率分布特性和自相关性,所提方法可较精确地评估风电并网系统充裕度及风电容量可信度。     

Considering wind speed correlation of WECS in reliability evaluation using the time-shifting technique

Wind power is a very useful renewable energy source that is attracting consideration around the world due to its non-exhaustive nature and its environmental and social benefits. The correlation between multiple wind sites has a significant impact on the reliability of power systems containing wind energy conversion systems (WECS). Conventional methods cannot be directly applied to evaluate the reliability of WECS in the absence of comprehensive modeling techniques that recognize the correlation of the wind speeds at different wind farm locations. This paper proposes a model for power system reliability assessment that can consider the wind speed correlation and preserve the statistical characteristics of wind speeds, such as the mean and deviation of the wind speed time series (WSTS). A time-shifting technique is used to produce a new WSTS for a given correlation between two wind sites. The optimal shifted time at the two sites is determined using a linear interpolation technique. The probability distributions of the generated power and the system reliability indices with different degrees of correlation between the two sites are compared using two reliability test systems, i.e. the RBTS and the IEEE-RTS. The results show that the proposed method is useful in evaluating the reliability of WECS with correlation between two wind sites.     

含大规模风光电源电力系统随机生产模拟

从电源规划的角度,利用随机生产模拟对新能源接入系统发电成本进行评估。根据风电与光伏电源的自身电源特性及其与电网的相互关系,提出风电与光伏电源参与随机生产模拟的原则。根据所提原则对应提出风功率损失期望及弃风比率2个评价指标。并利用IEEE-RTS测试系统验证了所提原则及方法的有效性与可行性。     

The impact of wind power plants on slow voltage variations in distribution networks

The connection of a wind power plant to a distribution network affects the voltage conditions in that part of the network. However, rather than assessing the worst possible cases regarding the voltage range, in this paper a more accurate methodology is presented for calculating the annual distribution network steady-state voltage profile and slow voltage variations after wind power plant connection. Commonly used simulation methods for the same problem based on probabilistic load flow analysis are only able to give probability distributions of the node voltages and network power flows. Here, a method is proposed which can also give chronological insight into the considered network state. This is realized through wind speed data modeling and generation using a Markov chain transition probability matrix and the Monte Carlo simulation method. The proposed methodology is tested using data from an existing distribution network in Croatia, a wind power plant planned for connection, and one year wind data measurements.     

基于改进马尔科夫链的风电功率时间序列模型

模拟风电功率时间序列在风电并网系统的规划和评估研究中具有重要意义,针对原始马尔科夫链在风电功率建模上无法保留其自相关性的不足,构建了一种基于改进马尔科夫链的风电功率时间序列模型。首先分析了风电功率的季节特性、日特性和波动特性;然后将风电功率数据按照不同月份及时段进行了细致划分,生成相应的状态转移概率矩阵;最后,对风电功率波动量的概率分布进行拟合,并叠加波动量,建立了基于改进马尔科夫链的风电功率时间序列模型。实例分析表明,本文所建新模型生成的风电功率序列能够保留历史序列自相关性,同时在一般统计参数、概率密度分布和自相关性三方面的准确性也优于已有模型。     

基于局部极差变化率的风电功率波动定量刻画

针对目前风电功率波动性研究中缺乏对其时序演进特征定量刻画的问题,对风电场实测功率数据样本进行分析,提出一种基于局部极差变化率的风电功率持续波动状态的识别方法,提取风电若干个持续出力状态以描述风电功率的持续波动特征。以用来衡量局部极差变化率的幅值和相角为模型输入量,建立灰色多目标决策模型,通过兼顾幅值的变化和相角的变化以寻找模型次优解的方法挖掘出具有代表性的幅值和相角,进而定义表征波动的量即波动系数,并以此来量化风电功率在某一时间段内的波动。给出了使用波动系数修正现行风电场预测预报考核指标的方法。     

Probabilistic power quality indices for electric grids with increased penetration level of wind power generation

In future electric grids it is expected that the share of power produced by renewable energy systems will increase to supply large deficits in power demand. Wind energy is one of the most important sources of renewable energy generation systems. With increased penetration level of wind energy conversion systems in modern electric grid, the quality of power will inevitably be affected. Power quality (PQ) indices are used to quantify the quality of the power. They serve as the basis for comparing negative impacts of different disturbances on power networks. Previous research of PQ indices with electric grids including wind energy sources was mainly based on fixed wind speed. Therefore, the PQ indices were calculated as instantaneous values that do not reflect the overall power quality impact of the grid connected wind energy systems. The main objective of this paper is to propose new probabilistic PQ indices for electric grids including wind energy systems. The proposed PQ indices combine both the probabilistic nature of wind speed using discrete Markov analysis and the electric grid behavior. The main PQ indices proposed are those concerning harmonics, flicker, and voltage sag. The developed indices are suitable for electric grids that include high penetration level of wind based power sources. The method used is general and can be applied to other power quality indices or power system performance indices.     

基于动态谐波回归的超短期风电功率预测

准确的风电功率预测对电力系统的安全稳定运行十分重要。从风功率统计特征出发,提出进行风电功率超短期预测的动态谐波回归方法。首先利用风电功率与不同高度风速的三次函数关系构建回归模型;然后采用自回归移动平均 模 型（auto regressive integrated moving average model,ARIMA）对回归的残差建模来充分利用风电功率时间序列的历史信息;最后针对风电功率的日季节性特点,引入傅里叶级数形成最终预测模型。经风电场实际数据计算验证表明,该方法有效弥补了ARIMA方法和回归方法的不足,减小了风电预测均方根误差（root mean squared error,RMSE）,提高了风电预测精度。通过和持续法、ARIMA 2种现有预测方法比较,验证了所提模型具有更高的预测精度,说明该方法具有一定的实际应用价值。     

Energy storage system-based power control for grid-connected wind power farm

Wind power is characterized as intermittent with stochastic fluctuations, which can result in deviation of grid frequency and voltage when the wind power ratio is high enough. These effects have a definite impact on stability and power quality of grid operation. This paper proposes an energy storage system (ESS) based power control for a grid-connected wind power system to improve power quality and stability of the power system. Vanadium redox flow battery (VRB), as an environmentally-friendly battery provided with many advantages, is employed in the ESS. A dynamic mathematical model of VRB is built by using an equivalent circuit, and its charging and discharging characteristics are analyzed. The VRB’s stable voltage is available in a wide range (around 20–80% state of charge), which is suitable for utilization of a single-stage AC/DC converter in the VRB-based ESS. With a proposed energy storage control method, VRB-based ESS is added at the exit of the grid-connected wind farm to filter fluctuations of wind power, which ensures that smooth power will be injected into the grid and which improves power quality of the power system. Simulations and experiments are carried out to verify the proposed power control method for these grid-connected wind power systems. The grid-connected wind farm with VRB-based ESS, wind speed (characterized as gust and stochastic wind), and wind turbines are modeled in simulations. Simulation results show that the grid-injected active power from the wind farm is effectively smoothed, and reactive power support can be provided for the grid by the designed VRB-based ESS. Experimental verification is achieved with a low power bench, where a RT-LAB real-time simulation platform and a direct torque controlled induction motor simulate a real wind turbine and a wind speed model is built in the RT-LAB real-time simulation platform. The experimental results verify the proposed scheme through demonstrating a stable and smooth power flow injected into the grid though the wind power fluctuated.     

含风电场的发输电组合系统可靠性评估

为评价风电并网对系统可靠性的影响,考虑了风速的时序性和自相关性,建立了风速的自回归移动平均模型,采用东台风电场实际监测的风速历史数据,并结合机组、线路和变压器等状态模型,建立了基于序贯MonteCarlo仿真方法的风电场发输电可靠性模型,对含风电场的发输电组合系统进行可靠性评估,同时采用了最优切负荷的方法,尽量减小切负荷量,并给出了具体的算法流程。算例采用IEEE-RTS测试系统,利用可靠性指标缺负荷概率LOLP、电量不足期望EENS和电力不足期望LOLE等反映风电并网对系统可靠性的影响。通过对仿真结果的比较和分析,可看出风电机组的接入对提高发输电组合系统的可靠性具有一定的积极作用。     

基于等值频率响应模型的海上油田群电网风电穿透功率极限分析方法

提出了一种基于等值频率响应模型的海上油田群电网风电穿透功率极限分析方法。首先,构建了海上油田群电网的平均系统频率响应(ASFR)等值模型,基于单、分轴燃机确定电网原动机-调速器参数,根据t location-scale分布和差分变换方法刻画风电出力波动概率分布;然后,基于电网可承受的最大风电波动对应累积概率指标和最严重故障后电网最大频率变化指标,提出了考虑稳态频率偏差约束和暂态频率稳定约束的风电穿透功率极限分析方法;最后,以某实际海上油田群电网为例验证了所提方法的准确性与有效性,结果表明,原动机-调速器响应速度和暂态频率稳定约束对于海上油田群电网的风电穿透功率极限具有关键影响。     

基于互转换Ornstein-Uhlenbeck过程的风速仿真模型及应用

准确的风速仿真是研究含风能发电系统的重要且基础步骤.为此,提出基于互转换Ornstein-Uhlenbeck过程的风速仿真模型,该模型可产生任意时间步长的仿真风速样本,将其与时序Monte Carlo模拟法结合,提出仿真时间步长可变的含风能发电系统充裕度评估方法.然后,采用某观测站的实测风速样本验证所提模型,结果表明,...     

基于解析法的风电场可靠性模型

采用频率和持续时间法的思想,通过风电场的风速数据和风电机组特性参数求得风电场各个状态的概率、频率、转移率。在马尔科夫链法的基础上,充分考虑风电场出力的随机性特点和风机的强迫停运率,建立了基于解析法的风电场可靠性模型。该模型采用聚集的思想,将风电场建模成一个类似于多状态的常规机组,大幅减少了风电场输出功率状态数。在RBTS测试系统中加入10台相同的V80-2 MW风机进行仿真,利用可靠性指标缺负荷期望(LOLE)、缺负荷频率(LOLF)、缺负荷持续时间(LOLD)反映风电场对系统可靠性的影响。通过对仿真结果的比较分析,证明了所提模型的可行性和有效性。     

大规模风电接入系统的备用优化分配研究

计及风电功率预测误差和负荷预测误差,定义3类发电负荷曲线以及对应的系统变量,考虑因不确定性波动导致发电机组增减出力后的系统安全约束,构建含风电系统节能调度的最优备用分配模型。以IEEE-RTS24系统为例,通过与常规含风电系统节能调度计算结果比较,证明了所建模型的有效性,并分析了在不同目标函数权重下,发电机组功率分配和备用分配对系统耗煤量的影响。     

基于EEMD和LS-SVM模型的风电功率短期预测方法

原始风速信号具有的间歇波动性特征给风电场的功率预测带来了挑战,采用集合经验模态分解(EEMD)法将原始风速信号分解为频域稳定的子序列,有效地提高了预测精度,避免了传统经验模态分解(EMD)存在的模态混叠现象。提出一种改进型果蝇优化算法(FOA),将风速子序列重构参数和最小二乘支持向量机(LS-SVM)参数作为优化目标建立风速预测模型,扩大了参数搜索范围,提高了优化收敛速度;通过风速-风功率转化关系可以求得风电场的功率值。实验结果验证了所提方法相比于EMD和LS-SVM预测方法具有更高的预测精度。     

基于灰色－辨识模型的风电功率短期预测

为了准确预测风电机组的输出功率,针对实际风场,给出一种基于灰色GM(1,1)模型和辨识模型的风电功率预测建模方法,采用残差修正的方法对风速进行预测,得出准确的风速预测序列。同时为了提高风电功率预测的精度,引入FIR-MA迭代辨识模型,从分段函数的角度对风电场实际风速-风电功率曲线进行拟合,取得合适的FIR-MA模型。利用该模型对额定容量为850 kW的风电机组进行建模,采用平均绝对误差和均方根误差,以及单点误差作为评价指标,与风电场的实测数据进行比较分析。仿真结果表明,基于灰色-辨识模型的风电机组输出功率预测方法是有效和实用的,该模型能够很好地预测风电机组的实时输出功率,从而提高风电场输出功率预测的精确性。     

不同风速概率分布参数下的风力抽水储能分析

介绍建立的风力抽水储能模型,推导了风能转换为电能后抽取的水质量关系式.在此基础上综合考虑风电场风速的随机特性,提出用Monte Carlo仿真产生每小时的风速抽样来计算风能特征指标.最后通过算例分析利用该方法计算了风速概率分布参数对抽水量的影响.结果表明,风力抽水储能是可行的.