基于 Super SAB神经网络算法的主变压器故障诊断模型

人工神经网络(ANN)由于其高度的非线性映射能力在电力系统模式识别及非线性优化领域有着广泛深入的应用研究.本文将Super SAB神经网络算法应用于主变压器溶解气体故障诊断(DGA),通过与带动量因子的标准BP算法、Bo1d Driver算法比较,验证SuperSAB算法在故障模式识别中具有更好的学习效率与泛化能力,故障诊断的准确度高于传统分析方法,表明其在变电设备状态诊断中具有良好的应用前景.     

Artificial Immune Network Classification Algorithm for Fault Diagnosis of Power Transformer

Dissolved gas analysis is an effective method for the early detection of incipient fault in power transformers. To improve the capability of interpreting the result of dissolved gas analysis, an artificial immune network classification algorithm (AINC), inspired by the natural immune system that is able to respond to an almost unlimited multitude of foreign pathogens, is proposed in this paper. The immune network system describes the complex interaction of antibodies and antigens in virtue of some immune mechanisms, such as immune learning, immune memory, etc. AINC mimics these adaptive learning and defense mechanisms to respond to fault samples of power transformers. Consequently, AINC can find a limited number of antibodies representing all fault samples distributing structures and features, which helps to realize dynamic classification. This proposed AINC algorithm has been tested by many real fault samples, and its results are compared with those of IEC ratio codes and artificial neural networks, which indicates that the proposed approach has remarkable diagnosis accuracy, and with it multiple incipient faults can be classified effectively     

Selection of Most Relevant Input Parameters Using Principle Component Analysis for Extreme Learning Machine Based Power Transformer Fault Diagnosis Model

The dissolved gas-in-oil analysis is a prevailing methodology being extensively utilized to diagnose incipient faults in oil-immersed power transformers. However distinct approaches have been implemented to find out dissolved gas analysis (DGA) results, they may sometimes fail to diagnose precisely. The incipient fault identification accuracy of various artificial intelligence (AI)-based methodology is assorted with change of input parameters. Thus, selection of input variable to an AI model is major research area. In this paper, principle component analysis algorithm using RapidMiner is applied to 360 experimental datasets, imitated in lab to identify most pertinent input variables for incipient fault classification. Thereafter, multi-class Extreme Learning Machine (ELM) technique is implemented to classify the incipient faults of power transformer and its performance is compared with artificial neural network, gene expression programming, fuzzy-logic, and support vector machine. The compared result shows that ELM provides better diagnosis results up to 100% accuracy at proposed input variable in short of time period which is helpful in on-line condition monitoring.     

Comparison of Test Methods for Characterization of High-Leakage Two-Winding Transformers

High-leakage transformers play an important role in power conversion. On-chip, silicon-integrated, and printed-wiring-board-integrated transformers are currently being developed for signal and power applications in high-frequency power supplies. High-power transformers have been developed for the inductive charging of electric-vehicle batteries. These transformers can feature relatively high winding resistances, core loss, and leakage inductances compared with conventional transformers. The conventional open-circuit and short-circuit test calculations have limited use for characterizing these relatively high parasitic transformers. In this paper, the differential (series-opposing) and cumulative (series-aiding) series-coupling tests are developed and applied for the accurate characterization of the transformer resistive and inductive elements. In the series-coupling tests, the various resistive and inductive components simply sum together, making transformer characterization more direct and accurate than attempting to interpret the standard open- and short-circuit (SOS) tests. The SOS tests are extended to provide a more accurate three-measurement approach, and improved correlation is demonstrated between experimental results obtained from the series-coupling tests and the extended open- and short-circuit tests. These results are validated by predictions from finite-element analysis. A sensitivity analysis is carried out to determine the sensitivity of the various test methods to measurement error. The cumulative test is highlighted as being particularly sensitive to measurement error compared with the other test methods.     

A differential relay for power transformers using intelligent tools

This paper presents alternative approaches using artificial neural networks (ANNs) for the protection of power transformers. A complete protection scheme was implemented. An ANN subroutine was used to discriminate internal faults from other situations, replacing the traditional Fourier method for harmonic restraint. In addition, a routine for reconstruction of saturated current signals based on recurrent ANNs is also proposed. The proposed methods were extensively tested and then compared to the traditional differential protection algorithm, showing promising results. The application of the ANN tools is a new and important stage in the differential relay analysis methodology for power transformer protection.     

基于特征判定系数的CSO-SVM的电力变压器振动信号故障诊断

变压器带电故障诊断对于保证电力变压器安全平稳运行具有重要的意义。针对变压器工作环境复杂且单一参数表征变压器故障类型不全面的问题,提出一种基于自适应噪声完备集合经验模态分解（CEEMDAN）和特征熵权法（EWM）进行故障诊断的方法。通过相关系数加权峭度（CCWK）原则筛选CEEMDAN分量并重构信号,实现在剔除冗余分量的同时,提升变压器振动信号特征的表征能力;利用EWM构建特征判定系数（FDC）实现单一数据诊断变压器故障类型;通过主成分分析法（PCA）减小混合域特征尺度,采用鸡群优化算法（CSO）优化支持向量机（SVM）模型进行故障诊断。对某变电站110kV三相油浸式变压器进行分析,结果表明与概率神经网络（PNN）和SVM等变压器故障诊断方法相比,文中方法在提前定性故障类型的同时,能够进一步提高变压器故障诊断准确率与效率。     

抑制变压器地磁感应电流的电容隔直装置安装位置优化

磁暴在电网中引起的地磁感应电流(GIC)导致变压器直流偏磁,对电力系统产生不利影响。在中性点安装电容隔直装置是治理变压器直流偏磁的常用方法,但由于GIC在电网中的流通路径复杂,在某个变压器安装隔直装置时,如不经充分考虑,往往会引起相邻变压器直流偏磁更加严重,因此文中研究电容隔直装置的安装位置优化问题。考虑自耦变压器接线和电网拓扑结构,引入变压器有效GIC来描述地磁暴对变压器的影响,分析了隔直装置的安装对变压器有效GIC分布造成的变化,提出了优化方法,在保证所有变压器的有效GIC小于允许限值的条件下,以隔直装置的安装数量最少作为优化的目标,并应用遗传算法求解。以甘肃主电网为例,构建包含47个变电站、101个节点的GIC等效模型,根据约束条件与优化目标,计算了隔直装置安装的数量和位置,并与未经优化的治理方案进行比较,验证了所提方法的可行性和优越性。     

Improved operation of power transformer protection using artificialneural network

This paper suggests the possibility of improving digital power transformer protection. The establishment of inrush in power transformers is becoming unreliable in existing numerical protection. An artificial neural network (ANN) was applied to inrush detection. The saturation of protective current transformers (CT) cannot be totally eliminated despite proper dimensioning. ANN was used for the reconstruction of distorted secondary CT currents due to saturation. In both cases, an ANN was included in the protection algorithm as an extension of the existing methods, which improved the reliability of the protection operation. The paper presents the digital protection algorithm completed in this way and the laboratory equipment by means of which experimental results were obtained. The results confirm faster and more reliable recognition of transformer inrush, as well as satisfactory reconstruction of the distorted secondary CT currents     

Mathematical models for current, voltage, and coupling capacitorvoltage transformers

This paper reviews a number of mathematical models used to represent the nonlinear behavior of the magnetic core of instrument transformers. Models of instrument transformers using these core representations are presented. The transient response of the instrument transformer is compared to actual test results recorded in the laboratory. The paper provides practical guidelines as to which of the physical elements of instrument transformers are important to model during transient studies and which elements could be ignored without sacrificing the accuracy of the simulation results. The electromagnetic transients program (EMTP) data files used to generate the models are also provided in the appendix to help new EMTP users model instrument transformers for evaluation of high-speed protective relaying systems     

Diagnostics of the power oil-filled transformer equipment of thermal power plants

Problems concerning improvement of the diagnostics efficiency of the electrical facilities and functioning of the generation and distribution systems through the examples of the power oil-filled transformers, as the responsible elements referring to the electrical part of thermal power plants (TPP), were considered. Research activity is based on the fuzzy logic system allowing working both with statistical and expert information presented in the form of knowledge accumulated during operation of the power oil-filled transformer facilities. The diagnostic algorithm for various types of transformers, with the use of the intellectual estimation model of its thermal state on the basis of the key diagnostic parameters and fuzzy inference hierarchy, was developed. Criteria for taking measures allowing preventing emergencies in the electric power systems were developed. The fuzzy hierarchical model for the state assessment of the power oil-filled transformers of 110 kV, possessing high degree of credibility and setting quite strict requirements to the limits of variables of the equipment diagnostic parameters, was developed. The most frequent defects of the transformer standard elements, related with the disturbance of the isolation properties and instrumentation operation, were revealed after model testing on the real object. Presented results may be used both for the express diagnostics of the transformers state without disconnection from the power line and for more detailed analysis of the defects causes on the basis of the advanced list of the diagnostic parameters; information on those parameters may be received only after complete or partial disconnection.     

数字量输出电子式电压互感器的高精度在线校验方法

针对目前电子式互感器校验方式存在的需要线路停电、操作复杂等问题,文中提出了一种数字量输出电子式电压互感器的在线校验系统。利用设计的菱形升降结构实现标准互感器与被测母线的连接,构建了一种可对母线电压信号进行在线测量的标准电压在线测量单元,从而实现了运行中电压互感器的在线校验。文中详细分析了同步方式、软件算法等因素对系统准确度的影响,并提出了降低误差的措施。研制的校验系统通过检定并在现场应用,准确度优于0.05级。     

智能全站仪ATR实测三维精度分析

简述了智能型全站仪TCA2003的自动目标识别（ATR）功能，以琅琊山抽水蓄能电站大坝变形监测网应用ATR功能三维测量实例，基于对其测量成果精度的评定，分析了ATR三维测量的精度，运用数理统计原理探讨了ATR测量的可靠性，对比了常规的人工测量与ATR测量的工效，给出了有益的结论，为同类工程的应用以及拓展ATR应用领域提供了技术依据。     

Thermal Analysis of Power Transformer Using an Improved Dynamic Thermal Equivalent Circuit Model

Abstract

—To insure the health of power transformers insulation, its heat distribution in different operating conditions must be determined. In extended thermal equivalent circuits (TECs), less attention has been paid to losses and temperature distribution of different parts such as core, tank and metallic regions of transformer. Most previous works, hot-spot temperature of winding and top-oil were considered by available no-load and on-load losses of transformer. This paper extends the TEC to be able to evaluate fully the different operating conditions of transformer. Estimation of losses of different elements of transformer as heat generating source and thermal parameters of transformer are introduced as two major challenges of the proposed model. At this end, an appropriate three-dimensional finite element method is used to estimate the losses of different components of power transformer. In addition, to identify the transformer thermal parameters, a genetic algorithm is used by searching in chosen measurements. Finally, the results obtained by the TEC are compared with IEEE-Annex G method and measured temperature and high accuracy of the proposed model in temperature estimation in different parts of transformer are shown.     

基于PSO⁃ELM融合动态加权AdaBoost的变压器故障诊断方法

油中溶解气体分析(DGA)是电力变压器故障诊断的一种常用方法。然而,基于DGA的传统方法不能建立精确的数学模型,单一智能诊断方法在实际应用中存在分类能力不足等问题,导致诊断结果准确率较低。文中提出一种基于PSO⁃ELM融合动态加权AdaBoost的变压器故障诊断方法。首先利用PSO算法对ELM模型进行优化,得到PSO⁃ELM模型,然后利用多分类AdaBoost算法对PSO⁃ELM模型进行重复迭代,得到带有不同权重的弱分类器,通过计算各分类器对测试样本的分错率,不断调整每个弱分类器的权重系数,最终利用所有弱分类器加权投票对监测数据进行故障诊断。实验结果表明,相比于BP神经网络、ELM、支持向量机(SVM),文中算法的诊断准确率分别提高16.02%、9.78%、5.62%。     

电力变压器关键尺寸视觉检测方法及其缩比模型验证

电力变压器的关键尺寸测量是其装配制造、运输安装过程中重要一环,现有测量方法操作繁琐且效率较低。为此,本文提出了一种适用于110 kV油浸式电力变压器关键尺寸视觉检测方法,该方法利用YOLO目标检测算法、Grabcut图像分割算法实现关键组部件的智能检测与分割,然后基于双目立体视觉原理实现套管相间距离及变压器最大截面外形尺寸等关键尺寸的测量。本文搭建了110 kV油浸式电力变压器外观缩比模型,试验分析了拍摄距离和角度等因素对电力变压器关键尺寸视觉检测的影响。结果表明,本文基于缩比模型试验实现了在不同拍摄距离与角度下变压器关键尺寸的视觉检测,验证了该方法的有效性。本文方法可以为110 kV油浸式电力变压器现场尺寸测量提供参考。     

基于改进人工鱼群算法的电网可用传输能力计算究

利用改进人工鱼群算法构造了可用传输能力问题的优化模型。首先采用潮流校验法解决了人工鱼群算法的初始值敏感问题;然后引入遗传算法,解决了人工鱼群算法的早熟问题;最后运用IEEE 30节点算例系统验证了所提算法的准确性和有效性,同时分析了人工鱼群算法在电网可用传输能力研究中尚需解决的问题。     

Fuzzy learning vector quantization networks for power transformercondition assessment

To improve the assessment capability of power transformers, this paper proposes a new intelligent decision support system based on fuzzy learning vector quantization (HVQ) networks. In constructing the system, a fuzzy-based classifier is designed to divide the historical data for dissolved gas analysis (DGA) into various categories with different levels of gas attributes. For each category of gas attributes, a learning vector quantization (LVQ) network is trained to be responsible for the classification of the potential faults due to insulation deterioration. The assessment approach has been tested on the DGA data from Taiwan Power Company (TPC) and compared with the previous fuzzy diagnosis system and the existing multi-layered backpropagation based artificial neural networks (BPANN) methods. Remarkable classification accuracy and far less training efforts of the proposed approach are achieved in this paper     

最小化能量损耗的配电网络动态无功优化算法

提出了一种完整的配电网络非线性混合整数动态无功优化模型及数学优化算法.该模型以全天能量损耗最小为目标,其约束条件不仅包括了各个时间断面的等式和不等式约束,而且还包括了有载调压变压器分接头和可投切并联电容器组的动作次数约束.为此,提出采用非线性原对偶内点法内嵌变量离散化罚函数的方法求解该模型,所推导出的高阶修正方程具有特殊的分块矩阵结构,可以充分利用稀疏处理技巧以提高算法的速度及对配电网络规模的适应性.最后以广州鹿鸣网配电系统作为算例,分析不同动作次数限制对配电网络电容器最优投切结果的影响,验证了该算法的正确性和有效性以及在限制控制设备动作次数方面取得的成功.     

Phase shifting transformer model for direct approach power flow studies

This proposal is intended to extend the field of application of an extremely efficient power flow algorithm used in radial and weakly meshed grids, the so-called Direct Approach (DA) method. In this work the method is broadened with the possibility of handling shunt admittances, transformers with taps, and phase shifting transformers. While the integration of the two former elements in the DA solver is quite straightforward, the use of phase shifting transformers is far from obvious due to their inherent non-symmetrical admittance matrix. Thus, a model for phase shifting transformers is proposed in this contribution, which allows the use of the DA method in grids that include such devices. A set of case studies is conducted in the contexts of a balanced industrial grid and a standard testbed to demonstrate the validity of the proposal.