电网故障诊断完全解析模型的解集评价与最优解求取

电网故障过程中保护和断路器的动作及其动作告警存在不确定性,致使基于电网保护规则解析的故障诊断模型存在多解问题,如何快速有效地求取最优解是电网故障诊断解析模型实用化的关键。针对此问题,提出一种求取电网故障诊断解析模型最优解的方法,通过解析保护和断路器动作的不确定性和告警信号的不确定性,构建完全解析模型的解集评价指标,将故障诊断问题转换为求取完全解析模型最优解的数学规划问题,然后利用保护和断路器的拒动、误动概率以及告警信号的误报和漏报概率等先验知识,采用随机优化算法求取最优解,并从最优解导出诊断结论。故障诊断算例和实际故障案例的诊断验证了所提出的完全解析模型求解方法的有效性以及相应故障诊断方法的容错性和实用性。     

基于故障群组合优化的变电站故障诊断

为处理好变电站故障诊断问题中故障源和故障征兆间因果关系的不确定性，提出了基于故障群组合优化的故障诊断方法。依据变电站运行方式和保护配置建立保护配合的Petri网模型，用该模型的状态转移方程获取所有的故障群-征兆群对子，创建故障群组合作为已知故障征兆诊断解的适应度函数，并用遗传算法搜索出最优诊断解。实验证明：该方法不仅能有效地辨识出故障源，而且能提供保护装置拒动、断路器拒动、信号传输错误等信息，对现场人员迅速确认故障设备非常有利。     

高压断路器合闸故障的快速查找方法

<正>1 问题的提出断路器是电力系统中重要的供电设备，断路器故障能否快速处理，恢复供电，关系到电力系统的稳定与可靠。近几年来，随着微机保护技术在断路器二次回路中的广泛应用，断路器合闸故障也呈现出一些新的特点，其故障查找、处理方法也与以往常规保护的合闸故障不尽相同。因查找方法不当，查找效率低，造成停电时间长，影响用能单位安全稳定生产。因此，对高压断路器合闸故障进行梳理、总结，实现对合闸故障的快速查找，是十分必要的。     

基于模拟进化理论的电力系统的故障诊断

本文提出了利用基于模拟分子进化（ＳｉｍｕｌａｔｅｄＭｏｌｅｃｕｌａｒＥｖｏｌｕｔｉｏｎ──ＳＭＥ）的优化方法诊断电力系统故障的新方法。电力系统故障诊断问题可表示为０－１整数规划问题，利用ＳＭＥ方法可一次得到几个可能的解，这对于诊断复杂的电力系统故障（尤其是有保护或断路器误动作的情况）具有实际意义。这种方法可以同时利用保护和断路器的信息，改进后也可以只利用断路器的信息，适用范围较广。此外，ＳＭＥ方法在适应能力方面优于现有的基于Ｂｏｌｔｚｍａｎｎ机的方法，可以处理任意多重故障，且适于用并行处理机实现。     

只利用断路器信息诊断电力系统故障的高级遗传算法

本文研究了基于断路器跳闸时序信息的电力系统故障诊断方法。主要包括三个方面的工作：①首次构造了只利用在各类保护（如主保护、第一后备保护和第二后备保护等等）动作时段内断路器跳闸的分段时序信息识别故障元件的０－１规划模型。②提出了应用无源信息识别故障区域的方法，有效地把故障诊断问题局限于某（些）个小的局部网络之中，大大加快了故障诊断的速度。③探讨了高级遗传算法（ＲｅｆｉｎｅｄＧｅｎｅｔｉｃＡｌｇｏｒｉｔｈｍ──ＲＧＡ）在电力系统故障诊断中的应用，这种方法可以有效地求得多个全局最优解，很适用于解决电力系统复杂故障情况下的诊断问题。经算例计算表明，所发展的故障诊断的数学模型是正确的，采用的基于高级遗传算法的故障诊断方法有在线应用的潜力。     

断路器越级故障的几种常见处理方法

针对断路器拒动或保护拒动可能引起越级故障的情况,介绍几种处理断路器越级故障的方法,通过实际案例分析,验证利用断路器失灵保护、增加远跳功能、利用主变压器后备保护、继电保护双重化配置等方法,可以很好地解决断路器越级故障.     

利用有限相量测量单元的故障识别与广域后备保护策略

针对传统后备保护整定困难、动作延时长、配合关系复杂等问题,提出了一种利用有限相量测量单元的故障识别与广域后备保护策略。首先,根据相量测量单元的安装位置划分后备保护区域;在发生故障时,利用电气量信息确定故障设备所在的后备保护区域,以期快速缩小可疑故障设备范围。其次,提出一种基于解析模型的故障识别方法和广域后备保护策略。当主保护未能成功切除故障时,这种解析模型可综合利用主保护动作警报、断路器动作警报、后备保护启动信息和功率方向继电器指向信息,利用禁忌算法（tabu search,TS）求解最优故障假说,并通过解析保护和断路器的动作逻辑判断拒动装置。之后,根据不同的拒动情况提出相应的后备保护策略以防止故障扩大。最后,以IEEE 39节点系统为例,对所提出的方法做了验证。仿真结果表明,所提出的方法是可行的,具有较强的容错能力。     

220 kV保护死区问题的探讨及改进

在继电保护的设计、施工、调试等实践工作中,有时因为疏忽导致保护死区的产生。该文简要分析了在现场接线中存在的几种常见的保护死区故障,即:各类断路器与断路器电流互感器之间的死区故障。简单介绍了几种保护死区故障产生的原因,针对这几个问题提出了几种解决保护死区的方法,并强调了在处理这些问题时应该引起高度重视的相关注意事项。     

电网故障诊断的一种完全解析模型

现代大规模互联电网提高了电力系统运行的稳定性和经济性,同时也给电网的故障诊断增加了难度。对电网保护配置和断路器动作规则进行深入分析,提出一种新的电网故障诊断模型——完全解析模型,并为基于此模型的故障诊断提出一种实用有效的求解方法。在完全解析模型中,故障假说（即可疑故障设备的故障情况）与保护、断路器的动作情况及拒动和误动情况一同表示成逻辑变量,而保护配置和断路器动作规则以逻辑方程组形式进行充分表达,逻辑方程组的每个解析解对应一个故障模式（设备故障状态和保护动作、断路器跳闸状态及拒动/误动情况）,是对故障情形的完整解释。本模型对故障诊断规则和保护、断路器的动作逻辑进行完全解析,并完整保留了故障设备、保护动作和断路器跳闸之间的耦合关系,克服了已有解析模型的缺陷。算例表明,保留完整耦合关系的完全解析模型可有效提高故障诊断的准确性和容错能力。     

220kV保护死区问题的探讨及改进

在继电保护的设计、施工、调试等实践工作中,有时因为疏忽导致保护死区的产生.该文简要分析了在现场接线中存在的几种常见的保护死区故障,即:各类断路器与断路器电流互感器之间的死区故障.简单介绍了几种保护死区故障产生的原因,针对这几个问题提出了几种解决保护死区的方法,并强调了在处理这些问题时应该引起高度重视的相关注意事项.     

Fault section estimation in power systems using a genetic algorithm

In this paper, a new method of fault section estimation in power systems using genetic algorithms (GAs) is presented. The main contributions of this paper include the following three aspects: (a) the fault section estimation problem is formulated as a 0–1 integer programming problem; (b) an efficient method to identify the faulty subnetworks is developed by using information from circuit breakers, and thus the fault diagnosis can be fulfilled in a very short time for large-scale power systems and can be implemented online; (c) a new method based upon a genetic algorithm is used to solve the fault section estimation problem, and the simulation results show that the GA based method can find multiple optimal solutions directly and efficiently in a single run, which is very suitable for complex fault diagnosis problems, especially for situations where protective relays and/or circuit breakers malfunction, because different combinations of fault sections and protective relay and/or circuit breaker malfunctions may give the same alarms.     

Fault Section Estimation of Power System Using Colored and Timed Petri Nets

A correct and rapid inference is required for practical use of expert system for fault section estimation in power systems. This paper proposes a new model-based expert system for fault section estimation using colored and timed Petri nets. The Petri nets simplify the inference procedure greatly because it can represent the causalities among faults of power system components, protective actions of relays and circuit breakers. They can also represent the operating states of these protective devices explicitly, or, more precisely, whether a relay has been actuated correctly or falsely, etc. In addition, time sequence information on actuated relays and tripped circuit breakers can be handled by using the colored and timed Petri nets. Therefore, the operating states of relays and circuit breakers can be inferred more correctly. Petri nets are suited for parallel processing so well that there is a possibility of reducing the inference time considerably. In the proposed system, the inference is executed by setting the initial states in the Petri nets using the information on the actuated relays and the tripped circuit breakers and then operating the Petri nets. There have been several test cases which used a prototype system on the Macintosh computer employing Object Common Lisp, and good results have been obtained.     

利用保护和断路器信息的电力系统故障诊断与不可观测的保护的状态识别的模型与 Tabu 搜索方法

根据元件故障与保护动作和断路器跳闸之间的逻辑关系,首先构造了电力系统故障诊断与不可观测的保护的状态识别（ＦＳＥ－ＳＩＵＰＲ）的一种新的０－１整数规划模型。该模型充分利用了断路器信息,在一些保护信息不完整的情况下也能正常工作。之后,提出了用Ｔａｂｕ搜索（ＴＳ）方法来求解这一问题。ＴＳ方法是近年来出现的用于求解组合优化问题的高效的启发式搜索技术。算例计算表明,所发展的ＦＳＥ－ＳＩＵＰＲ的数学模型是正确的,采用的基于ＴＳ的方法相当有效。     

A new neural networks approach to on-line fault section estimationusing information of protective relays and circuit breakers

This paper proposes a new neural network diagnostic system for online power system fault section estimation using information of relays and circuit breakers. This system has a similar profile of an expert system, but can be constructed much more easily from elemental samples. These samples associate fault section with its primary, local and/or remote protective relays and breakers. The diagnostic system can be applicable to the power system control center for single or multiple fault sections estimation, even in the cases of failure operation of relays and breakers, or error-existent data transmission. The proposed approach has been practically verified by testing on a model power system. The test results, although preliminary, suggest this system can be implemented by various electric utilities with relatively low customization effort     

Fault section estimation in power system using Hebb’s rule and continuous genetic algorithm

In this paper a new approach for fault section estimation (FSE) in electrical power system is presented. We propose a procedure to obtain objective function (required for fault section estimation) using the Hebb’s learning rule. The continuous genetic algorithm (CGA) optimization method is then employed to estimate the fault section making use of the objective function. The Hebb’s learning law used in this paper gives, linear algebraic equations, to represent the targets in terms of the status of relays and circuit breakers (CBs). This gives a simple objective function, which leads to reduction in time required by the CGA to identify fault section. The CGA gives an advantage of requiring less storage than binary genetic algorithm (GA). Also the CGA is inherently faster than binary GA.The proposed approach is tested on various systems, and is found to give correct results in all cases. Simulation results for two illustrations have been presented in this paper. The results show that the proposed approach can find the solution efficiently even in case of multiple faults or in case of failure of relays/circuit breakers. A comparison with artificial neural network (ANN) approach is also presented.     

基于贝叶斯网络的电网故障诊断

针对电网故障诊断中存在的信息具有不确定性的问题,依据元件故障、保护动作和断路器跳闸之间的内在逻辑关系,由Noisy-Or和Noisy-And节点组成贝叶斯网络和采用类似训练多层前馈神经网络的误差反传算法进行诊断模型的参数学习,分别建立了线路、变压器和母线的通用故障诊断模型;依据元件-保护-断路器间的关联关系,给出了元件诊断贝叶斯网络的自动生成方法,最后对各个元件的诊断网络进行推理,以获得元件的故障概率值。实例仿真表明了该诊断方法的可行性和有效性,无论简单故障或多重故障,并且存在保护和断路器拒动、误动的情况下,都能得到合理有效的诊断结果。     

基于小波神经网络的在线警报处理系统

利用小波变换在信号处理方面的时频分析能力和神经网络对任意非线性函数的普通的逼近能力，提出了一个基于小波神经网络的电力系统故障段辨别方法。故障诊断系统依据保护继电器和断路器的采样信息估计电力系统中故障段的位置。仿真结果显示，小波神经网络故障诊断系统能正确估计电力系统单一故障和多重故障的位置，即使在电力系统中存在保护继电器和断路器误动或拒动的情况下，小波神经网络也能给出合理的结果。测试结果表明，小波神经网络在电力系统警报处理系统中有良好的应用前景。     

电力系统故障诊断的量子粒子群优化算法

电力系统故障诊断是利用保护和断路器的动作信息来推断可能的故障位置.其中故障元件的识别是故障诊断实现的关键.文中应用量子粒子群优化算法研究故障元件的识别方法,先根据保护动作原理将故障诊断问题表示为0-1整数规划问题,然后用量子粒子群优化算法求解.与标准PSO算法和传统遗传算法比较,文中采用的量子粒子群优化算法具有稳定性高...     

A Modified Abductive Inference Model for Fault Section Estimation in Power Systems Using the Tabu Search Approach

A new approach to fault section estimation in power systems is presented based upon a modified abductive inference model and the tabu search method. The developed fault section estimation model can simultaneously take into account the operating reliabilities of protective relays and circuit breakers and the degree of correctness of received and nonreceived alarm data in a formal and systematic manner. In the modified abductive inference model, a criterion for describing the relative plausibility of different diagnosis hypotheses is available. Based on this criterion, the fault section estimation problem is then formulated as a 0-1 integer programming problem, and a tabu search (TS) approach is presented for solving the problem. A sample power system is served for demonstrating the correctness of the developed abductive inference based fault section estimation model and the computational efficiency of the TS-based method.