On-line monitoring of dissolved gas-in-oil with FTIR spectra

To overcome the disadvantages of conventional DGA (dissolved gas-in-oil)analysis using gas chromatography and other electrochemical sensors, initial researches werecompleted to realize on-line monitoring of dissolved gas-in-oil of power transformers using FTIR(Fourier Transform InfraRed) spectroscopy. Gas cell method is used to determine the characteristicabsorption peaks of each diagnostic gas; simple and novel devices and procedures were designed inorder to get measurable samples and spectra of mixed diagnostic gases with known concentration aretaken using long optical path gas cell. The range of wavelength is estimated to be 3.0-13.9 mum fromexperimental spectra data. Hence the corresponding sampling frequency range should be in 536-4288Hz and usable optical materials are suggested. It is concluded that a resolution of 10 cm^(-1) maywell satisfy the monitoring of all diagnostic gases and water content except hydrogen, and thelowest detection limit may be as low as 2Xl0^(-8) to acetylene with a 2.4-meter-long optical length.

Optimum Optical Length of the Gas Cell Used for Monitoring Gas-In-Oil with FTIR

Dedicated experiments are designed to collect the infrared spectra of dissolved gas-in-oil of power transformers. Spectra of diagnostic gases are collected by 3 different laboratorial FTIR spectrometers using 3 different gas cells with various sets of equipment parameters. A formula is deduced to calculate the shortest optical length to detect a specific concentration according to measurements on gases with known concentrations near to the minimum detection limit. Collected spectra and calculated results suggested that the optimum optical length of the gas cell should be 150 mm to realize on-line monitoring of diagnostic gases within the required concentration range. At the end, an economic novel design of the gas cell is proposed based on the optimum length.

Power Transformer Fault Diagnosis Using Fuzzy Reasoning Spiking Neural P Systems

This paper presents an intelligent technique to fault diagnosis of power transformers dissolved and free gas analysis (DGA). Fuzzy Reasoning Spiking neural P systems (FRSN P systems) as a membrane computing with distributed parallel computing model is powerful and suitable graphical approach model in fuzzy diagnosis knowledge. In a sense this feature is required for establishing the power transformers faults identifications and capturing knowledge implicitly during the learning stage, using linguistic variables, membership functions with “low”, “medium”, and “high” descriptions for each gas signature, and inference rule base. Membership functions are used to translate judgments into numerical expression by fuzzy numbers. The performance method is analyzed in terms for four gas ratio (IEC 60599) signature as input data of FRSN P systems. Test case results evaluate that the proposals method for power transformer fault diagnosis can significantly improve the diagnosis accuracy power transformer.

A Heuristic Approach to the Diagnosis of Transformer’s Insulating Oil

The transformer plays so important equipment in power system that engineers take more measures on the insulating oil of transformer by diagnosis. The dissolved gas analysis (DGA) is an effective technique for detecting incipient faults in oil-immersed power transformers. So the paper investigates the DGA methods, while employ the ANSI/IEEE C57.104 standards and the Key Gas diagnosis rules as base to develop a fast transformer fault diagnosis method in practice. I designed a report’s form which was so easy to understand that we can have accurate diagnosis what was up in the body of transformer by EXCEL programmed. The user only keys in the measured data of main gases including CO, H2, CH4, C2H2, C2H4, and C2H6 those gases were taken from ASTM D3612’s instruction. Then the diagnosis result was showed in texts and the plotted figures which were two figures to compare diagnosis the test’s figure with the reference figure of the Key Gas diagnosis rules that was taken the analysis of transformer fault from over past in power system. Last but not least, the proposal offers a simple, quick, and an accurate of diagnosis through human-machine interface. While which was been quickly, simply, and accurately proved on October 25th, 2012 Nan Cou E/S #4 ATr’s insulating oil of diagnosis.

A New Approach with Three Dimension Figure and ANSI/IEEE C57.104 Standard Rule Diagnoses Transformer’s Insulating Oil

The dissolved gas analysis (DGA) is an effective method for detecting incipient faults in immersed oil power transformers. In this paper, we investigate the DGA methods and employ the ANSI/IEEE C57.104 standards (guidelines for the interpretation of gases generated in oil-immersed transformers) and IEC Basic Gas Ratio method to design a heuristic power transformer fault diagnosis tool in practice. The proposed tool is implemented by a MATLAB program and it can provide users a transformer diagnosis result. The user keys in the data of H2, CH4, C2H2, C2H4, and C2H6 gases dissolved from the immersed oil transformer’s insulating oil measured by ASTM D3612. The analyzed results will be represented in texts and figures. The real measured data of the transformer oil were taken from Taiwan Power Company substations to verify the validation and accuracy of the developed diagnosis tool.

A Hybrid Diagnosis Approach on Transformer’s Insulating Oil

The immersed-oil power transformer is so vital equipment in power system that maintenance-engineers take more monitor from transformer’s insulating oil to diagnose what is condition of operation. Then the dissolved gas analysis (DGA) is known for an effective technique on detecting incipient faults in oil-immersed power transformers. In this paper, a practical method is presented which consists of the Roger & Dernenber Ratio Methods, the Linear SVM diagnosis, the Key Gas method and the Specification ANSI/IEEE C57.104 Standard. Thus, incipient faults in power immersed-oil transformers can be directly identified by a report’s form which is so easy understood that we can accurate of diagnosis transformer. The user only keys in the measured data of main gases such as H2, CH4, C2H2, C2H4, C2H6, and CO those gases were must decompose via ASTM-D3612. The diagnosis result was showed in texts. This paper was taken some data from Taiwan and Siemens Power Company to verify the program that was validation and accuracy of the transformer’s insulating oil diagnosis tool.

Fault Diagnosis of Power Transformer Based on Improved ACGAN Under Imbalanced Data

The imbalance of dissolved gas analysis(DGA)data will lead to over-fitting,weak generalization and poor recognition performance for fault diagnosis models based on deep learning.To handle this problem,a novel transformer fault diagnosis method based on improved auxiliary classifier generative adversarial network(ACGAN)under imbalanced data is proposed in this paper,which meets both the requirements of balancing DGA data and supplying accurate diagnosis results.The generator combines one-dimensional convolutional neural networks(1D-CNN)and long short-term memories(LSTM),which can deeply extract the features from DGA samples and be greatly beneficial to ACGAN’s data balancing and fault diagnosis.The discriminator adopts multilayer perceptron networks(MLP),which prevents the discriminator from losing important features of DGA data when the network is too complex and the number of layers is too large.The experimental results suggest that the presented approach can effectively improve the adverse effects of DGA data imbalance on the deep learning models,enhance fault diagnosis performance and supply desirable diagnosis accuracy up to 99.46%.Furthermore,the comparison results indicate the fault diagnosis performance of the proposed approach is superior to that of other conventional methods.Therefore,the method presented in this study has excellent and reliable fault diagnosis performance for various unbalanced datasets.In addition,the proposed approach can also solve the problems of insufficient and imbalanced fault data in other practical application fields.

电力变压器在实际应用中对于绝缘杂项的传感及检测特性研究

油浸式电力变压器内油中溶解气体的含量直接影响着高压绝缘装置的绝缘性能,实现变压器内油中溶解微量气体(H_(2),C_(2)H_(2),CO及CH_(4))的可靠检测对于高压绝缘装置及电网安全至关重要。在本研究中,对长期在电网中运行的油浸式电力变压器利用密度泛函理论研究了过渡金属Zn原子掺杂的S空位SnS_(2)单层(Zn-SnS_(2))敏感材料对4种油浸式变压器中油中溶解气体的传感吸附行为。结果表明:用于油浸式电力变压器油中溶解杂项的传感材料在线监测导带和价带结构得到了有效优化(由1.550 eV变为0.127eV),掺杂体系下不同带隙和电导率的吸附体系将有助于提高新型二维SnS_(2)基气敏材料对油浸式电力变压器油中溶解气体的选择性,为实现电力变压器油中溶解气体的可靠检测提供了一种新的思路。

基于油中溶解气体特征量筛选的变压器故障诊断方法

油中溶解气体分析对变压器故障预警及诊断具有重要意义。针对油中溶解气体特征量种类众多、故障关联特征分析不足等问题,文中以油浸式变压器为研究对象,提出了基于油中溶解气体特征量筛选的变压器故障诊断方法。首先,对油中溶解气体的原始特征量进行特征衍生,通过随机森林(random forest,RF)计算特征量对故障诊断的重要度,筛选得到最佳特征组合。其次,采用树结构概率密度估计(tree-structured parzen estimator,TPE)实现RF模型的参数寻优,并形成TPE-RF诊断模型。同时,结合多种评价指标,证明所提方法能够对变压器作出准确的故障诊断。最后,提出TreeSAHP模型分析特征量对各故障的重要度,优选出各故障关联的主要特征量,并根据变压器运行案例,探讨了该方法在电力行业现场应用中的适用性,验证了该方法的有效性。

电力变压器油中溶解气体离群值识别和数据重构

高质量传感数据是驱动新型电力系统数字化和智能化发展的基础,而由于传感器性能退化、传输中断或其他干扰因素,数据时常出现错误和异常值,造成数据利用率低等问题。针对在役电力变压器油中溶解气体在线监测数据,该文提出了基于COPOD、孤立森林(IForest)与Grubbs的联合方法提升油中溶解气体数据的价值。首先,通过COPOD和IForest筛选出包含离群点的数据集,再采用Grubbs对其进行检验,有效识别离群值。进一步地,采用掩码方式优化训练Transformer神经网络模型,填补空缺值重构油中溶解气体数据序列。在相同气体数据序列上,所提算法正确识别点数、正确识别离群点数和受试者工作特征曲线平均面积相比于传统K-近邻算法分别提升了3.5%、29.4%和5.0%。对于数据填补,对比双向缩放算法,填补后的数据与实际数据的方均根误差均值为7.29μL/L,平均绝对误差均值为2.7μL/L,性能分别提升了9.7%和9.2%,有效地提高了数据的质量和利用率。最后,通过11台500 kV变压器油中溶解气体数据分析,有力支撑了变压器状态评价和设备数字化管理。

基于光强调制型光热干涉法的绝缘油中乙炔传感

实时跟踪变压器油中乙炔气体对保障油浸式变压器稳定运行具有重要意义。新型光学光热干涉技术(photothermal interference,PTI)具有“零背景”、高灵敏度、无接触测量和抗电磁干扰等的优点,在变压器油中溶解气体分析领域具有巨大潜力。该文提出一种基于Herriott气室的光强调制型光热干涉技术的乙炔传感技术。首先,对单光束激发乙炔光热效应的过程进行仿真建模,研究光热效应引起的温度分布、脉冲激励动态响应和正弦激励的频率响应;进一步,开展参数优化实验。结果表明,增大激励光光强能有效增大光热相位,而提高调制频率不利于光热效应的充分激发;最后,开展模拟故障状态下的油中乙炔检测,PTI乙炔系统的检测灵敏度为0.15(m V·L)/μL,检测下限达到0.94μL/L,满足在运高电压等级变压器油中乙炔含量的监测需求,验证光热干涉法在油中溶解气体检测的可行性和潜力。

变压器故障诊断用油中溶解气体新特征参量

油中溶解气体分析（dissolved gas analysis,DGA）是现场电力变压器故障诊断最常用的方法。然而,油中溶解气体含量较容易受到变压器结构、容量、故障位置以及故障程度等因素的影响,从而降低了变压器故障诊断的可靠性。为了提升变压器故障诊断正确率,该文提出了基于支持向量机（support vector machie,SVM）和遗传算法（geneti calgorithm,GA）优选的DGA新特征参量。首先,以28个DGA比值为输入,建立了基于SVM的变压器故障诊断模型;其次,采用GA同时对SVM参数和DGA比值进行优化,得到9个优选DGA比值作为变压器故障诊断用新特征参量。对IEC TC 10故障数据库的诊断结果表明：DGA新特征参量的故障诊断正确率为84%,较常用的DGA含量和IEC比值的诊断正确率提高10%~25%;并且无论采用哪种特征参量,支持向量机的诊断结果均优于神经网络诊断模型。最后,采用DGA新特征参量对国内117组变压器的故障诊断正确率达到了87.18%,再次验证了该方法的有效性。

基于遗传算法进化小波神经网络的电力变压器故障诊断

在电力变压器故障诊断方法中,小波神经网络常用的反向传播算法存在着易陷入局部极小点和对初值要求较高的缺点,往往给故障诊断带来困难。文中提出了一种基于遗传算法进化小波神经网络的变压器故障诊断方法,用实数编码的遗传算法来代替人解决小波神经网络结构的选择和参数的设定。在整个学习过程中,网络的复杂度、收敛性和泛化能力得到了较好的综合。大量实例表明,该方法能有效地对电力变压器单故障和多故障样本进行分类,提高了诊断准确率。

基于灰关联熵的充油变压器故障诊断方法

油中溶解气体分析是目前发现变压器潜伏性故障的重要方法,鉴于用IEC推荐的三比值法中编码缺陷(编码超出码表)及变压器故障诊断的复杂性,文中详细阐述了如何将灰关联熵应用于变压器故障诊断。首先通过统计方法,选择典型油中气体作为参考列,并经反复调整,挖掘出油中气体所含故障信息,然后利用灰关联熵方法进行变压器故障类型诊断。该方法基于融合互补的思想,将灰关联分析方法与信息熵理论有机结合起来,克服了单一灰关联分析中易造成局部关联及信息损失等缺陷,尽可能多地包含变压器本体所含信息。实例分析结果表明,该方法具有较好的分类效果。

基于BP网络算法优化模糊Petri网的电力变压器故障诊断

为了提高电力变压器故障诊断的正确率,提出了一种基于BP网络算法优化模糊Petri网的电力变压器故障诊断方法。利用具有自学习、自适应能力的BP网络算法,在确定模糊Petri网的权值、阈值、可信度等网络参数初始值的前提下,实现模糊Petri网网络参数的优化。在模糊Petri网网络结构上,运用BP网络算法,对电力变压器DGA样本进行学习训练,使模糊Petri网网络参数逐步向真实值逼近。实例分析结果表明,该方法能够有效地诊断电力变压器中的单一故障和多重故障,提高故障诊断正确率,证明了方法的正确性和有效性。

油气分析诊断变压器故障方法的改进

针对油中溶解气体分析 (DGA)中原来采用比值法的某些不足 ,通过对大量数据的统计、分析 ,在变压器吊芯结果所确定典型故障分类的基础上 ,提出了比值范围法改进三比值变压器故障诊断。

基于改进小波神经网络算法的电力变压器故障诊断方法

大型电力变压器作为电力系统的枢纽设备,其运行可靠性直接关系到电力系统的安全与稳定。针对基于BP算法的小波神经网络存在收敛速度慢、搜索空间局部极小及易引起振荡等不足,本文以变压器油中溶解气体为分析对象,提出采用动量项和变学习率改进小波神经网络的变压器故障诊断算法。选择400组油中溶解气体含量作为小波神经网络训练及故障识别样本,对训练过程和仿真结果进行对比分析。实验结果表明:较之比值法,改进的小波神经网络故障诊断算法在故障识别准确率和收敛时间方面表现更优。

基于深度自编码网络的电力变压器故障诊断

基于深度自编码网络(DAEN),构建了分类深度自编码网络(CDAEN)模型。结合电力变压器在线监测油中溶解气体分析(DGA)数据,提出了基于CDAEN的变压器故障诊断方法。所提方法利用大量无标签样本进行预训练,优化模型参数,并利用少量有标签样本进行微调。实例分析表明,与基于反向传播神经网络(BPNN)、支持向量机(SVM)的故障诊断方法相比,所提方法的诊断正确率更高。

电力变压器故障诊断的人工免疫网络分类算法

变压器油中溶解气体分析是电力变压器绝缘故障诊断的重要方法。文中将人工免疫网络分类算法应用于电力变压器故障诊断,利用增加抗原、记忆抗体类别信息的人工免疫网络对故障样本进行学习,可以获取更好地表征故障样本特征的记忆抗体集,再用最邻近分类法对故障样本进行分类。经大量实例分析,并将其结果与IEC三比值法和BP神经网络等方法的结果相比较,表明该算法能有效地对电力变压器单故障和多故障样本进行分类,具有较高的诊断准确率。

基于云物元分析原理的电力变压器故障诊断方法研究

变压器油中溶解气体分析(DGA)是电力变压器故障诊断的重要方法。针对物元理论变压器故障诊断方法中,在建立故障模式物元模型时没有考虑边界值的不确定性的不足,首次在变压器故障诊断研究方面引入云模型,结合云模型的不确定推理特性以及物元理论能同时进行定性定量分析问题的优点,提出了一种基于云物元分析原理和DGA相结合的电力变压器故障诊断新方法。通过建立变压器故障诊断的云物元模型和计算特征云物元与标准云物元之间的关联函数,实现对变压器故障模式的有效识别。实例分析验证了方法的正确性和有效性。