基于层次分析法的油纸绝缘局部放电模式识别

通过对局部放电的模式识别可以了解放电类型及严重程度,并在此基础上确定维护方案。为了对局部放电进行识别,建立了油纸绝缘中的5种典型缺陷模型;运用K-W检验从相间局部放电(PRPD)统计算子中提取出分类能力最强的11个特征;基于提取的特征,在小样本训练集的前提下,利用层次分析法对典型放电模型进行识别,同时和同种情况下使用人工神经网络的识别效果进行了比较。实验结果表明,在小样本训练集下,运用层次分析法得到了较好的识别效果,正判率均大于85%,优于人工神经网络,这为小样本训练集的情况下局部放电的快速识别奠定了基础。     

PD source identification with novel discharge parameters usingcounterpropagation neural networks

Computer aided partial discharge (PD) source identification using different multidimensional discharge patterns is widely regarded as an important tool for insulation diagnosis. In this paper, a neural network (NN) approach to PD pattern classification is presented. The approach is based on applying variants of the counterpropagation NN architecture to the classification of PD patterns. These patterns are derived from physically related discharge parameters, different from those commonly used. It is shown that considerable improvements of the classification quality can be obtained when an extended counterpropagation network with a dynamically changing network topology is applied to patterns that employ the voltage difference between consecutive pulses instead of the phase of occurrence as the main discharge parameter. Furthermore, using a particular parameter vector that takes the correlation between consecutive discharges into account also allows to solve the rejection problem with this type of NN     

基于核典型相关分析融合的XLPE电缆局部放电模式识别

提出使用核典型相关分析方法提取XLPE电缆接头局部放电信号PRPD图谱特征信息,并使用K最近邻分类算法实现不同绝缘缺陷模式的高准确率识别。利用YJV-26/35 k V型电缆及其附件设计了4种典型绝缘缺陷,使用脉冲电流检测获取局部放电样本信息,绘制了PRPD图谱并应用于样本数据,研究不同特征向量下的识别效果,在适合维数最终获得较高识别正确率。相对于传统电力设备模式识别方法,不但可以有效反映信号非线性特征,并可以将多种特征进行有效融合,消除冗余特征。     

基于ARTNN的GIS绝缘故障识别新方法

为根据局部放电信号识别早期的GIS绝缘故障缺陷类型,提出了一种利用ART神经网络在线识别GIS绝缘故障类型的新方法。较之常用BP神经网络,该法训练时间短、所需样本少、权值稳定、不存在局部收敛,故更适于在线识别。网络的输入量为一个工频周期内局部放电脉冲重复率、主频率、阻尼系数、放电量、放电相位分布。利用5种GIS绝缘缺陷类型的实验所得数据对ART神经网络进行训练及验证,证明该法的缺陷类型正确识别率可>98%,在GIS绝缘故障类型的在线模式识别中具有广泛的前景。     

气体绝缘组合电器多局部放电源的检测与识别

介绍了基于宽带检测的气体绝缘组合电器(gas insulated switchgear,GIS)多局放源检测与识别技术。该技术由脉冲群快速分类和基于最小二乘支持向量机(least square-support vector machine,LS-SVM)的基于相位分布的局部放电谱图(phase resolved partial discharge,PRPD)识别2个主要模块组成。其中脉冲群快速分类模块由基于脉冲波形的时频参数提取和竞争学习网络无监督聚类实现,它将脉冲群分为若干个由相同波形特征脉冲组成的子脉冲群。PRPD放电谱图识别模块对各子脉冲群对应的PRPD谱图进行放电指纹提取,并使用GIS绝缘缺陷特征数据库训练得到的LS-SVM判别函数对各子脉冲群进行识别。仿真和试验结果均验证了该技术的可行性和实用性。     

变压器内单一局放PRPD谱图统计参量的主成分分析

设计了4类由变压器油纸绝缘缺陷引起的"单一局放模式"模型:气隙放电模型、针板放电模型、介质表面金属杂质放电模型、油隙放电模型。通过标准化试验方法得到了不同模型局部放电相位分布模式(phase resolved partial discharge,PRPD模式)的二维谱图,并对谱图进行分析得到25个局部放电统计参量。采用主成分分析对统计参量的有效性进行了分析,得到了25个统计参量对不同类型放电信息表达的新特征参量组,以及对不同放电类型识别的贡献率。     

Bootstrap方法在局部放电特征提取中的应用

介绍了用Bootstrap方法对局部放电特征的小样本原始数据集重复再采样 ,提取其特征参量的统计信息 ,建立绝缘缺陷特征隶属函数 ,并应用遗传编程分类方法对缺陷模型进行模式识别的方法。试验结果表明该方法有效、可行。     

基于遗传编程的绝缘内部局部放电缺陷模式识别

采用新的模拟进化技术——遗传编程，进行局部放电模式识别以区分不同的绝缘内部缺陷类型。制作了4种结构的人工缺陷模型以模拟发电机定子中典型的绝缘内部放电，从局部放电试验中获得二维和三维谱图特征，计算局放信号的矩特征值。首先用模糊方法将局部放电信号的矩特征表示为关于对象不确定知识的模糊特征，作为放电数据的预处理。再由遗传编程分类表达式进化生成局部放电缺陷类型判别函数，并采用递增式学习规则以提高最佳特征对局部放电缺陷分类的效果。另外，将Bootstrap统计模拟技术与遗传编程结合，以克服从小样本数据中进行知识获取的“瓶颈”。人工缺陷模型试验数据的测试结果表明，该方法在局部放电缺陷类型识别中得到了良好的识别效果。     

油浸绝缘纸损伤与局部放电相位分布模式演化的关系

文章对油浸绝缘纸损伤过程中的表面形貌、表面粗糙度、表面电导率和气隙气体与局部放电相位分布(PRPD)模式演化的关系进行了研究。根据PRPD模式变化将放电损伤过程划分为五个阶段,利用显微镜、扫描电镜、原子力显微镜、高阻仪和气体测量装置,对比分析了每一阶段的表面损伤情况及气隙气体变化。研究结果表明:五个损伤阶段中,电负性气体含量交替下降上升,气隙内的主要放电形式在亚辉光(或辉光)放电和脉冲放电之间交替转换;亚辉光放电的PRPD模式主要由生成的弱电负性气体自身性质决定;脉冲放电的PRPD模式主要由表面损伤决定;脉冲放电的PRPD模式均拥有不同程度的"兔耳"特征和周期特征,"兔耳"特征强弱主要由表面电导率决定,周期特征强弱主要由表面陷阱密度决定;表面粗糙度不改变PRPD模式形状,但随着其值下降放电量呈上升趋势。     

PD monitor system for power generators

This paper introduces a computerized PD monitoring system for generators and presents the experimental and numerical study of discharge pattern recognition methods. The system has specially designed transducers, data acquisition unit and software, and can obtain statistical as well as individual discharge information. In order to validate the performance of the system, experiments were done in the laboratory, using elaborately designed models that can generate various types of discharges. Feature extraction of the gathered data and neural network (NN) classification of the acquired discharge patterns were studied. The results showed that the surface fitting method is able to extract features from statistical data of discharges, and that NN is a potential classifier in practical applications     

Wavelet analysis for classification of multi-source PD patterns

Multi-resolution signal decomposition (MSD) technique of wavelet transforms has interesting properties of capturing the embedded horizontal, vertical and diagonal variations within an image in a separable form. This feature was exploited to identify individual partial discharge (PD) sources present in multi-source PD patterns, usually encountered during practical PD measurements. Employing the Daubechies wavelet, features were extracted from the third level decomposed and reconstructed horizontal and vertical component images. These features were found to contain the necessary discriminating information corresponding to the individual PD sources. Suitability of these extracted features for classification was further verified using a radial basis function neural network (NN). Successful recognition was achieved, even when the constituent sources produced partially and fully overlapping patterns, thus demonstrating the applicability of the proposed novel approach for the task of multi-source PD classification     

几种特征选择方法在局部放电模式识别中的应用

局部放电模式识别的输入特征量选择是非常关键的步骤。针对油纸绝缘中5种典型局部放电类型,从其相间局部放电（PRPD）谱图中提取出31个统计算子。分别运用K-W检验、类内类间距离比、顺序前进法以及遗传算法等4种方法对这些算子进行了选择优化。分别用这些选取的特征量组合作为输入向量,通过BP神经网络这个统一的模式识别技术来比较研究这4种特征选择方法,结果表明,顺序前进法和遗传算法由于考虑了特征量之间的相关性,所选择的特征量优于另外2种方法。     

Application of genetic algorithms to pattern recognition of defectsin GIS

A computerized pattern recognition system based on the analysis of phase resolved partial discharge (PRPD) measurements, and utilizing genetic algorithms, is presented. The recognition system was trained to distinguish between basic types of defects appearing in gas-insulated system (GIS), such as voids in spacers, moving metallic particles, protrusions on electrodes, and floating electrodes. The classification of defects is based on 60 measurement parameters extracted from PRPD patterns. Classification of defects appearing in GIS installations is performed using the Bayes classifier combined with genetic algorithms and is compared to the performance of the other classifiers, including minimal-distance, percent score and polynomial classifiers. Tests with a reference database of more than 600 individual measurements collected during laboratory experiments gave satisfactory results of the classification process     

Trends in partial discharge pattern classification: a survey

Partial discharge (PD) detection, measurement and classification constitute an important tool for quality assessment of insulation systems utilized in HV power apparatus and cables. The patterns obtained with PD detectors contain characteristic features of the source/class of the respective partial discharge process involved. The recognition of the source from the data represents the classification stage. Usually, this stage consists of a two-step procedure, i.e., extraction of feature vector from the data followed by classification/recognition of the corresponding source. The various techniques available for achieving the foregoing task are examined and analyzed; while limited success has been achieved in the identification of simple PD sources, recognition and classification of complex PD patterns associated with practical insulating systems continue to pose appreciable difficulty.     

Partial discharge recognition system for current transformer using neural network and 2D wavelet transform

In this paper, a neural network and two‐dimensional (2D) wavelet transform are applied to recognize partial discharge (PD) patterns on current transformers (CTs). To avoid the discrepancy between simulated results and real experimental data, we adopted seven cast‐resin CTs that were purposely fabricated with various insulation defects as the PD patterns collected samples to actually emulate the various defects incurred often during their production. All measurements are taken in a shielded lab; the commercial TE571 PD detector is adopted to measure PD patterns to ensure the reliability of the PD signals. Next, we extract the patterns' features via a 2D wavelet transform and use the features as the training set of a backpropagation neural network (BNN) to construct the recognition system for CTs' PD patterns. Finally, we add random noises to the measured PD signals to emulate the field diagnosis under a high‐noise environment. The study results indicate that, under a simulated noise magnitude of 30 pC, the recognition rate of the proposed system still can reach around 80%, signifying a great potential in applying the proposed recognition system in field measurements in the future. © 2011 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

基于卷积神经网络的高压电缆局部放电模式识别

由高压电缆不同类型缺陷诱发的局部放电(PD)的识别难度较大,尤其是某些相似度较高的电缆绝缘缺陷类型难以区分。提出了一种基于卷积神经网络(CNN)的高压电缆PD模式识别方法,研究了不同网络层数、不同激活函数以及不同池化方式对识别效果的影响,并与传统的支持向量机(SVM)和反向传播神经网络(BPNN)算法进行了对比。结果表明,相比SVM和BPNN,CNN的总体识别精度分别提高了3.71%和4.06%,且能较好地识别具有高相似度的电缆缺陷类型。     

XLPE电力电缆附件局部放电测量与分析

为研究XLPE电力电缆附件现场常见典型缺陷的放电特征,在3根电缆实体上分别设计制作了中间接头尖刺、主绝缘划伤和终端头应力锥错位3种放电模型,建立了基于PDBase的局部放电测量分析系统。对比研究了3种典型缺陷的局部放电特征,分析了放电次数相位分布谱图H_n(φ)、放电最大幅值相位分布谱图H_(qmax)(φ)和放电幅值分布谱图H(q)3种统计特征。试验结果表明,不同缺陷类型其放电发展过程不尽相同,呈现的PRPD谱图、单个脉冲波形、相位分布趋势及统计特征区别明显;而同一缺陷在相同条件下其放电特征呈现出相似规律且重复性好;这些特征为进一步开展电缆附件放电机理研究及放电类型的模式识别提供了有力的试验依据。     

基于GK模糊聚类和LS-SVC的GIS局部放电类型识别

局部放电可以反映气体绝缘组合电器(Gas Insulated Switchgear,GIS)内部的绝缘缺陷,对正确识别GIS的放电类型具有重要意义。在GIS重症监护系统研究平台上人工设置4种GIS的典型缺陷。基于4种缺陷不同电压等级下的局部放电样本数据,提取局部放电灰度图像的分析特性作为识别特征量。同时考虑到现场干扰对局部放电信号的影响,利用GK模糊聚类算法对分形特征量进一步处理,以提取隔离干扰后的分析特征量。最后设计了基于LS-SVC的局部放电模式识别器。试验结果表明所提方法能有效识别GIS放电类型,比人工神经网络方法具有识别率高、稳定性好的优点。     

基于统计参数的自适应网络推理系统在局部放电缺陷识别中的应用

高压电力设备在发生绝缘劣化的早期,内部会出现局部放电现象,笔者依据检测得到的局放信号,提出了采用基于统计参数的自适应网络推理系统进行绝缘缺陷模式识别的方法。自适应网络推理系统是神经网络和模糊逻辑的结合,通过模糊逻辑进行识别系统建模,利用神经网络训练系统参数。设计并实验了4种绝缘缺陷模型,对多周期的局放信号进行相位分布及幅值分布统计,提取表征局放特性的统计参数,总结了不同缺陷模型局放特征的区别。实际的检测结果表明,经过训练后的局放缺陷识别系统,能够有效地对各种缺陷的样本数据进行分类,达到良好的识别效果。     

基于深度学习和多模型融合的局部放电模式识别方法

局部放电与电力设备的绝缘状态息息相关,准确识别局部放电类型对于保障电网运行具有重要意义。文中提出一种基于深度学习和多模型融合的局部放电模式识别方法。首先,设计并搭建开关柜内4类典型局部放电缺陷模型,采集局部放电相位分布(phase resolved partial discharge,PRPD)图谱并建立样本集;其次,搭建基于迁移学习的深度残差网络,对局部放电缺陷进行识别;最后,利用Sugeno模糊积分将深度残差网络(deep residual net￣work,DRN)模型与传统识别模型进行融合。实验结果表明：迁移学习模型相比于无迁移学习模型有着更好的更新能力和泛化性能;融合模型与单一模型相比具有更高的识别准确率。所提方法能够准确识别局部放电缺陷类型,对于电力设备的运维检修具有一定的参考价值。