基于PMU测量数据的改进型戴维南等值参数递推估计算法

研究快速获得电力系统戴维南等值参数,实现对指定节点的电压在线监测,提出一种利用微分变量反应参数变化的改进型戴维南等值参数递推估计算法,并推导出参数计算的表达式。根据电网节点电压电流相量特性,基于PMU测量数据计算出电网侧等值阻抗,将其作为该算法的启动初值。具有结构简单,计算量小的特点。经IEEE14节点系统算例仿真验证,能够快速有效的在线实时跟踪戴维南等值参数,并可广泛应用在实际电网中。     

适应大扰动的电力系统戴维南等值参数跟踪算法

针对现有的戴维南等值参数跟踪算法需要假设2个计算时步之间戴维南等值参数不变、不能适应等值系统内部发生扰动等问题,提出了一种可以适应系统内部出现大扰动的戴维南等值参数跟踪算法。该算法利用本地测量信息,主要解决了现有算法不能适应等值系统内部发生大扰动时戴维南等值参数计算问题,跟踪算法利用故障前的戴维南等值参数计算出故障时刻的等值电压源幅值,然后以故障前一时刻的等值电阻和故障时刻的等值电压源幅值作为初值,计算出故障时刻的其他戴维南等值参数,从而再求出故障后各个时刻的戴维南等值参数。该算法的优势在于仅利用负荷母线局部电气量就可以快速、准确、实时地求取戴维南等值参数,而且也适用于等值系统内部发生三相短路的情况,同时该算法具有计算结果误差小的优点。最后,通过IEEE 30和IEEE 39节点标准测试系统验证了本算法的准确性。     

基于网损等值负荷模型的改进直流最优潮流算法

交流最优潮流因需要处理大量非线性约束,求解效率不高,且在求解含有小阻抗支路和重负荷特征的病态电力系统时,容易出现不收敛的情况;而直流最优潮流计算速度快,可以处理病态电力系统,但是计算精度较低。基于此,文中提出了基于网损等值负荷模型的改进直流最优潮流算法。该算法首先采用网损等值负荷等效替代线路网损,有效地提高了原有直流模型的计算精度;然后使用简化原—对偶内点法进行求解,提高了算法的计算效率。通过对IEEE 30节点、IEEE 300节点、Polish 2736节点和Polish 3120节点系统的测试,证明了所提算法不仅有着较高的计算精度和效率,而且在不依赖于交流潮流解的基础上对于病态电力系统具有较强的处理能力。     

配电网无功优化逐层筛选寻优算法

利用配电网络具有树状层次结构的特点,提出了一种逐层筛选寻优的配电网无功优化算法.文中将辐射网逐层分解为随电压线性变化的等值负荷,并推导了等值负荷参数递推计算的方法;对含电容器、变压器分接头的控制策略空间也进行逐层分解,将每一个子策略映射到一组等值负荷参数,通过淘汰每一层无功电压负荷参数相似而局部有功网损较大的子策略,有效压缩了策略筛选空间,实现了自底向上的逐层筛选寻优.通过选取合适的参数,能保证算法在实现快速计算的同时具备全局寻优能力.另外,该算法优化结果具有确定性,不存在初值选取和收敛性问题.文中给出了实际配电网18节点系统和IEEE 69节点系统的算例仿真,并与传统遗传算法在计算速度和优化效果方面进行了对比,说明该算法适用于含离散变量的大规模辐射网无功优化问题的在线计算.     

采用可变等值注入法的等值潮流算法

根据网络分析理论提出了一种等值注入可变的系统等值模型，并由其原理给出了采用可变等值注入法的电力系统等值潮流算法。其特点是：等值计算即使是在非基态情况下仍具有较高的精度，并且对外部系统运行条件的变化有较好的响应能力。通过ＩＥＥＥ标准试验系统进行了实际计算与分析。计算表明：算法的原理正确，较之常规的等值方法，计算更准确，收敛性更好。     

带有复杂分支子馈线的配电系统可靠性评估

采用网络等值法对带有复杂分支子馈线的配电系统进行可靠性评估，并提出一种基于新的分层结构的等值算法。该算法通过将网络元件和分支子馈线组合成节点和线路2种集合元件，根据配电网实际拓扑结构，采用层状的数据结构描述节点集合元件和线路集合元件之间的拓扑关系，从而避免直接描述网络元件之间的复杂关系。并采用宽度优先搜索算法对其进行顺向和逆向遍历求解系统的可靠性指标。实例计算验证了该算法的有效性。     

REI等值法用于多节点配电系统短路电流计算的研究

对扩展的径向等值不变性等值网络应用局部节点等值方程进行了研究,提出了将ＲＥＩ等值法用于多节点配电系统短路电流计算,将外部系统中的电源等值联接地边界节点上,将边界节点归入保留节点。将原有源节点变成无源节点。简化了多节点配置电系统短路电流计算的数学模型,使短路电流计算程序更简便,实用,计算精度明显提高。     

复杂中压配电网的可靠性研究

针对复杂中压配电网的特点,提出了一种复杂中压配电网络可靠性评估快速算法—化简分块算法。该评估算法利用网络等值,将各分支馈线等效为等值元件,再利用邻接矩阵将简化后的系统划分为若干块,然后以块为单位进行故障分析,计算中采用基于故障扩散的搜索方法确定故障节点类型,从而计算每个负荷点以及馈线和系统的可靠性指标。利用该算法可快速地实现网络的可靠性评估,通过算例证实了其高效性和工程实用性。     

考虑系统内部扰动的戴维南等值参数跟踪方法

提出一种新型的戴维南等值参数计算方法。该算法能够解决传统戴维南等值参数求解方法中存在的等值系统内部发生扰动时,传统算法无法求解戴维南等值参数的问题及参数漂移问题。通过对传统算法的分析,提出基于泰勒展开式的Taylor算法,综合考虑系统中的非线性因素并建立校正机制,使得算法在复杂系统运行情况下表现出良好的准确性与稳定性。在3机9节点系统与IEEE 39节点系统中的仿真证明了所提算法的有效性。     

基于戴维南等值的电压稳定评估方法

从等值模型和参数辨识方法两方面分析了戴维南等值方法在电压稳定分析中的问题。首先从系统节点电压方程出发,验证基于极限点处戴维南等值阻抗与负荷阻抗相等这一性质构造的稳定裕度指标具有较好的实用性,而此方法的关键在于系统各时刻等值参数的跟踪。其次比较了现有参数跟踪算法,分析了含遗忘因子最小二乘递推算法应用于此时出现数值稳定性问题的根本原因,根据估计误差提出了改进算法。最后分别以2节点系统和10节点系统两个仿真算例验证了该算法的有效性。     

A modified linear algorithm for power system contingency analysis

This paper introduces an algorithm for fast linear contingency analysis which computes linearized system line real power flow changes due to system outages. It implements the relation between bus injected real powers and bus phase angles, taking into account the second-order changes in the relation due to the system contingencies. The effects of these changes greatly improve the DC load flow results over similar methods used previously for this type of problem.     

修正节点接入导纳潮流算法

针对牛顿–拉夫逊法初值选取不当和PQ分解法受网络RX比例影响造成的计算不收敛问题，提出了一种修正节点接入导纳的潮流算法。理论证明该算法是正确的。算例分析表明，该方法收敛可靠、计算速度快，同时不受RX比例与初值选取的影响。     

Real-Time Decoupled Equivalent Models for Static Security Analysis

This paper presents some new decoupled external system equivalent models for use in system security evaluation. Three different models i.e., decoupled Tie-line power flow model, decoupled Boundary bus Impedance model and decoupled Impedance-Admittance model are developed based on P-? and Q-V relations of fast decoupled load-flow formulation. The structure component of the models are then identified by processing real time data using Kalman filtering technique. Then the operating components of the decoupled equivalents are obtained from two new approaches. The performance of the various models with regard to contingency analysis has been presented using simulated data on a 25 bus test system. Persistency analysis of different input-output data is given to determine apriori choice of models. Also the bias conditions are derived to ascertain the accuracy of the models.     

发电联合转移分布因子及快速静态安全校核算法

发电转移分布因子(GSDF)和广义发电分布因子(GGDF)都是常用的灵敏度分析及安全评估工具。文中指出了GSDF和GGDF的缺陷，尤其对GGDF的普遍适用性提出质疑。提出了发电联合转移分布因子(GJSDF)和负荷模增发电联合转移分布因子(GJSDF-LIP)的概念，并导出若干有用的性质。针对电力市场环境下购电交易计划新模型和流畅调度算法的特点，将GJSDF和GJSDF-LIP成功地应用于静态安全校核算法中，极大地提高了求解速度。算例表明，该算法效率很高，并且易于编程实现。     

一种基于ANN技术的预想事故自动选择方法

本文提出了一种基于人工神经元网络（ＡＮＮ）技术的预想事故自动选择方法，利用快速解耦潮流计算的迭代一次法（ＦＤＬＦ１）构造了分别对应于支路有功潮流越限的有功功率性能指标ＰＩｐ 和对应于节点电压模值和发电机无功出力越限的电压－ 无功性能指标ＰＩＶＱ。用ＢＰ算法训练多层感知型神经元网络以求得对应于不同运行状态和不同网络拓扑结构的性能指标。通过算例分析，并与其它经典方法的比较，本方法在计算精度和速度方面非常令人满意，而且具有较高的事故捕获率。     

Dynamic Load Flow Technique for Power System Simulators

This paper presents a dynamic load flow technique for power system simulators. The approach presented can be applied to a system when there is a net accelerating or decelerating power in the system. It possesses all the attractive features and basic structure of the normal fast decoupled method, and it does not depend on the selected reference bus. The modifications to the standard fast decoupled power flow algorithm are derived, and power system simulator implementation considerations are discussed. The algorithm has been tested on standard IEEE test cases and on a metropolitan utility 559 bus model. A numerical example of the above algorithm on the 559 test case is presented which demonstrates convergence characteristics and computation requirements of the proposed algorithm. These results are compared with those of normal fast decoupled load flow.     

二重预想事故排序的快速计算

在线安全分析中 ,二重预想事故由于其后果的严重性已受到越来越多的关注。在本文中 ,提出仅考虑由于一重预想事故引发的继发性二重预想事故 ,即不考虑由于偶然原因引起的两预想事故同时发生的情况 ,在进行计算时 ,采用了完全边界法和自适应边界法相结合的方法 ,并提出了包含电压、无功与有功的综合性能指标 ,从而大大提高了预想事故筛选的速度 ,并且预想事故筛选过程同时进行了一重预想事故筛选和二重预想事故筛选 ,为运行人员提供了更多有用的信息。     

Pre-screening of single contingencies causing network topologychanges

A multilevel contingency screening method is introduced to process large sets of candidate power system contingencies. The idea is to apply a series of increasingly more accurate and time-consuming algorithms so that each higher screening level processes a reduced set of candidates. A practical implementation of the method uses a fast prescreening step whose algorithm is based upon the local network solution concept. The method is designed to process contingencies whose effects can be localized, e.g. branch outages and bus section faults. The accuracy of the new algorithm is controlled via an automatically calculated small network adequacy ratio while its reliability is enhanced by the use of different contingency severity measures     

A fast technique for outage studies in power system planning and operation

Security monitoring is a fundamental aspect of the operation of power systems. One of the basic tools that dispatchers use to assess the system security is contingency analysis. This program simulates the state of the system after the sudden loss of an element of production or transmission. The DC load flow used in the determination of the performance index (PI) yields only the real power loading of circuits, and ignores the effects of the bus voltage. Therefore, we present an AC active power model in this paper where the effect of the voltage magnitudes is considered. The changes of the coefficient matrix model due to the line outage are directly determined without reconstruction of that matrix or redetermination of its inverse. Then, the postoutage angle is obtained directly. A derivation of the PI is presented, including only the overloaded lines as a result of the fast contingency screening approach proposed in the paper. Real and practical values of the line power limits required for PI calculations are considered. Furthermore, a simple method for the determination of some matrices included in the PI formulation is introduced in this paper to reduce the computation time compared with that for the Wollenberg-Mikolinnas algorithm. The proposed approach is applied to an IEEE 14-node system and the ranking table is obtained. The proposed method was found to be excellent.     

A methodology for probabilistic simultaneous transfer capabilityanalysis

This paper presents a methodology for simultaneous transfer capability analysis based on a probabilistic approach. All areas in a large scale power system are divided into three groups: (a) study area, (b) transfer participating areas, and (c) external areas which have no direct transactions or they have fixed transactions with the study area. A performance index based contingency selection procedure is applied within the study and transfer participating areas to rank those contingencies which will affect simultaneous transfer capability. The contingency ranking order is utilized by a variation of the Wind Chime diagram to selected contingencies which are then evaluated by an optimal power flow algorithm. Subsequently, the probability distribution of simultaneous transfer capability is computed based on the electric load, circuit and unit outage Markov models. The 24×3 bus IEEE RTS is utilized to evaluate the proposed method. The performance of the proposed method is also demonstrated on an actual large scale system (2182 bus, 8 area system)