变压器故障油色谱诊断技术研究

发布时间：2018-04-04 18:07

本文选题：变压器　切入点：油色谱　出处：《吉林大学》2017年硕士论文

【摘要】：随着中国经济的飞速发展,电网的规模日益扩大,电力变压器作为电网中的枢纽设备,它承担着对电能二次分配和相互转换的繁重任务,在电网中有着举足轻重的地位。为提早发现变压器的潜在的隐患,降低变压器发生故障而带来的经济损失,变压器故障油色谱诊断技术的提升就显得十分的迫切。目前,国内外对变压器故障类型分析有许多好的方法和路径。本文从研究国内外变压器故障油色谱诊断法的应用现状及发展趋势入手,对该项诊断技术的原理和技术特点进行介绍,并对变压器油中气体特征的产生和溶解情况进行具体分析。电力变压器发生故障的几个主要类型,它主要包括过热性和放电性故障、局部受潮,并对上述三种故障类型分别进行了介绍。目前国内外普遍应用的几种电力变压器故障油色谱诊断技术,例如:气体组分谱图技术、油中溶解气体分析用专家系统技术、人工神经网络技术、三比值法、四比值法等分别进行了介绍,分析了这些诊断技术方法及其原理与优、劣势。重点介绍了变压器故障的灰色关联诊断技术和主因子诊断技术。利用灰色关联诊断技术所具有良好的诊断性能,可以将待诊断数据与标准数据类有机排比,并得到关联度。越高的关联度,其数值也就越大。通过标准数据进行灰度比对,可以有效提升变压器故障范围的判断。同时采用电力变压器故障主因子诊断技术,就可以将全部数据特征进行难度简化,并取得原始变量所提供的大量信息,从而提高了变量数据诊断的准确率。在此基础上提出主因子和灰色关联的组合诊断技术,并对变压器故障展开了组合诊断分析,进而提出组合诊断的模型。最后使用主因子与灰色关联的组合诊断技术对本地区三个典型案例进行了实验数据分析。统计出相对的故障类型结果之后,和变压器实施的停电检测结果比对基本符合。实践证明,本文提出的组合诊断技术能有效地提高故障诊断的正确率和范围,对提高电力变压器安全、稳定运行,有一定的理论意义与实用价值。
[Abstract]:With the rapid development of Chinese economy, the scale of power grid is expanding day by day. Power transformer, as the hub equipment in the power network, undertakes the heavy task of secondary distribution and mutual conversion of electric energy, and plays an important role in the power grid.In order to find out the potential hidden trouble of transformer and reduce the economic loss caused by transformer fault, it is very urgent to improve the technology of transformer fault oil chromatographic diagnosis.At present, there are many good methods and paths for transformer fault type analysis at home and abroad.In this paper, the application and development trend of oil chromatographic diagnosis method for transformer fault are studied, and the principle and characteristics of the diagnosis technique are introduced.The generation and dissolution of gas in transformer oil are analyzed in detail.There are several main types of power transformer faults, which mainly include overheating and discharge faults, partial dampness, and the above three types of faults are introduced respectively.At present, several kinds of oil chromatographic diagnosis techniques for power transformer faults are widely used at home and abroad, such as gas component spectrum, expert system technology for analysis of dissolved gas in oil, artificial neural network, three-ratio method, etc.The four-ratio method is introduced, and these diagnostic techniques, their principles, advantages and disadvantages are analyzed.The grey correlation diagnosis technology and the main factor diagnosis technology of transformer fault are introduced in detail.The grey correlation diagnosis technique has good diagnostic performance and can be compared with the standard data and the correlation degree can be obtained.The higher the correlation, the greater the value.By comparing the standard data with gray level, the range of transformer fault can be improved effectively.At the same time, the main factor diagnosis technology of power transformer fault can simplify the difficulty of all the data features and obtain a lot of information provided by the original variables, thus improving the accuracy of variable data diagnosis.On this basis, the combined diagnosis technology of principal factor and grey correlation is put forward, and the combined diagnosis analysis of transformer fault is carried out, and the combined diagnosis model is put forward.Finally, the experimental data of three typical cases in this area are analyzed by combining the principal factor and grey correlation.After the results of relative fault types are calculated, the results of outage detection of transformers are basically in accordance with each other.It has been proved by practice that the combined diagnosis technology proposed in this paper can effectively improve the accuracy and range of fault diagnosis, and has certain theoretical significance and practical value for improving the safety and stability of power transformer operation.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM407

【参考文献】