Return voltage method is a non-destructive electrical diagnosis technique for assessing of oil-paper insulation condition of large-scale power transformer. According to the existing research and application, dominant time constant is the most commonly used characteristic quantity, but it is difficult to define dominant time constant accurately. Moreover,it is difficult to separate ageing and moisture on solid insulation.The study is not deep on correlation between initial slope and insulation condition, which has influenced the accuracy of result obtained by the method. On the basis of the fact, this project studies the several key issues of the method. Firstly we will investigate on the polarization mechanism of oil-paper insulation system, revealing the correlation between polarization spectrum, initial slope and insulation condition were experimentally studied, making clear the change law of corresponding characteristic quantity with various insulation defects,such as ageing and moisture.simulate its polarization properties using equivalent circuit model, then establishe a mathematical-physical model for analysing insulation conditions; adopt the improved intelligence optimization algorithm to identify equivalent circuit parameters of dielectric response accurately. Secondly, characteristic quantity and polarization spectrum was caculated by the identifid circuit parameters, the feasibility of parameters identification are verified by comparing measured and simulated polarization spectrum. Thirdly, exploring for new ideas determines dominant time constant; studying change law of circuit parameters with variation of insulation condition by experiment; simulating change law of characteristic quantity with variation of identified parameters. Lastly, on the basis of simulation and experiments, the project will put forward a synthetic diagnosis method combining many features of dielectric response for insulation condition of transformer,such as dominant time constant, initial slope, polarization,conductivity and so on. The result of project will perfect basis theory of return voltage method, which has not only the important theoretical value, but also a good extensive practical application prospect.
回复电压方法是一种诊断大型电力变压器的油纸绝缘状态的无损电气诊断方法,它能现场评估变压器固体绝缘的状态。现有研究主要利用主时间常数特征量,但缺乏准确界定主时间常数的方法,而且该特征量难以区分变压器固体绝缘的老化和受潮,这影响了该方法诊断的准确性。基于此,本课题深入研究该方法存在的若干关键问题:揭示变压器固体绝缘的受潮和老化对其介质响应特性及回复电压特征量的影响机理,明确不同绝缘缺陷引起相应特征量的变化规律;利用等值电路模型模拟其极化特性,建立分析绝缘老化状态的物理数学模型,应用改进的智能优化算法,辨识等值电路参数;计算新的特征量和极化谱,仿真极化谱的变化规律,提出确定主时间常数的新方法;建立利用主时间常数、初始斜率等多个介质响应特征量区分变压器固体绝缘老化和受潮的综合评估模型,并提出诊断方法,提高诊断的可靠性。本项目的研究成果将完善回复电压诊断方法的基础理论,为工程应用提供理论基础。
回复电压方法是一种能现场评估变压器固体绝缘状态的无损电气诊断方法,但该方法缺乏准确确定主时间常数的方法,在利用主时间常数评估变压器固体绝缘的水分含量时结果偏高;难以利用现有特征量区分变压器固体绝缘的老化和受潮等,影响了该方法诊断的准确性。因此,揭示变压器固体绝缘的受潮和老化对回复电压特征量的影响机理及其变化规律,建立基于多个介质响应特征量综合评估油纸绝缘老化的模型和分析变压器绝缘状态的方法具有重要的学术研究价值和实际应用价值。. 本项目的研究难点是特征量的有效测量及在评估变压器固体绝缘老化和受潮中的应用。为了准确确定主时间常数,解决评估过程中数据量和信息量偏少的问题,提出了基于回复电压特征量的油纸绝缘等值电路的数学建模及参数辨识研究,并提出了改进的数学模型,该模型能减少测量点,缩短测量时间。为了区分绝缘老化和受潮对回复电压特征量的影响,定义了回复电压初始斜率峰值比特征量Sr1和Sr2,Sr1为充电1000s、放电1s的回复电压初始斜率峰值比,Sr2为充放电比为2时的初始斜率峰值比。研究表明这两个特征量对老化和受潮的敏感程度不同,适宜于区分老化和受潮;在此基础上,项目还研究了频域介电谱特征量的变化规律,提出了利用积分频域谱分析变压器油纸绝缘状态的新方法,并将其用于区分油纸绝缘的老化和受潮。. 项目取得的成果有:1、建立利用回复电压峰值、峰值时间以及初始斜率特征量综合求解介质响应等值电路参数的改进数学模型;2、提出了准确确定主时间常数的新方法和利用极化谱及其它介电特征量综合诊断绝缘状态的方法;3、定义了新的绝缘老化特征量,提出了利用不同充放电时间下的初始斜率及其与回复电压峰值比值区分绝缘老化和受潮的方法。4、提出了利用主时间常数,初始斜率与值比、低频介损和直流电导率综合分析和诊断变压器油纸绝缘状态的新方法。本项目的研究成果将完善回复电压诊断方法的基础理论,为工程应用提供理论基础。
{{i.achievement_title}}
数据更新时间:2023-05-31
带有滑动摩擦摆支座的500 kV变压器地震响应
城市轨道交通直流自耦变压器牵引供电系统故障保护研究
特高压冲击电压分压器线性度测量方法比较
A Fast Algorithm for Computing Dominance Classes
混合动力汽车电池性能影响因素分析与试验探究
基于电路分析法和恢复电压响应的油纸绝缘老化诊断方法
换流变压器油纸绝缘局部放电特性与机理研究及绝缘状态评估
基于介电响应技术的油纸绝缘状态评估方法研究
变压器绝缘纸板的纳米改性及其对复合电压下油纸绝缘电场分布均化作用的研究