Analysis of Statistical Methods for Assessment of Power Transformer Frequency Response Measurements

Analysis of Statistical Methods for Assessment of Power Transformer Frequency Response Measurements
复制标题

电力变压器频率响应测量评估统计方法分析

DOI:
10.1109/tpwrd.2020.2987205
复制
发表时间:
2021
影响因子:
4.4
通讯作者:
S. Miyazaki
S. Miyazaki
中科院分区:
工程技术2区
文献类型:
--
作者:
M. Tahir;S. Tenbohlen;S. Miyazaki

文献摘要

参考文献

被引文献

相似文献

频率响应分析 (FRA) 是一种广泛接受的诊断方法,用于评估电力变压器内有源部件的机械完整性。 FRA 领域的最新研究已经实现了 IEC 和 IEEE 标准中描述的测量程序的标准化。然而,FRA结果的解读仍然仅限于该领域专家的分析。使用数字索引是 FRA 解释引入的方法之一,它量化了 FRA 签名的差异。文献中提出了许多指数。然而,目前还没有全面的比较研究来评估这些数值指标对不同故障的敏感性。本次贡献就是朝这个方向的努力,将所有指数收集在一个平台上,根据它们的特点进行比较和分类,公布更合适的指数作为标准指数。本文定义了适当指标的标准,例如单调性、线性、灵敏度、数据点依赖性和指标比率。使用来自实验装置、模制变压器和实际电力变压器案例的 10 个案例研究来进行指数评估。本文的结果表明,一些指数在性能方面脱颖而出,可以被认为是能够令人满意地计算 FRA 签名之间变化的最合适指数。研究结果还为 FRA 研究中实施的数值指数标准化过程提供了有用的事实和指导。
Frequency response analysis (FRA) is a widely accepted diagnostic method to assess the mechanical integrity of the active parts within a power transformer. Recent studies in the field of FRA have led to the standardization of its measurement procedure as described in the IEC and IEEE standards. However, the interpretation of FRA results is still limited to the analysis of the experts in the field. Using numerical indices is one of the methods introduced for the FRA interpretation, which quantifies the differences in FRA signatures. There are numerous indices proposed in the literature. However, there is no comprehensive comparison study to assess the sensitivity of these numerical indices against different faults. This contribution is an effort in this direction, which collects all indices on a single platform, compares and categories them based on their characteristics, to announce more appropriate indices as the standard ones. This paper defines the criteria for an appropriate index such as monotonicity, linearity, sensitivity, data-points dependency and index ratio. The evaluation of the indices is performed using 10 case studies from an experimental setup, molded transformer, and real power transformer cases. The results of this paper reveal that some indices stand out from the others in terms of performance and can be considered as the most appropriate ones that can satisfactorily calculate the changes between FRA signatures. The results also provide useful facts and guidelines for the standardization process of numerical indices for their implementation in the FRA studies.
DOI: --
发表时间: 2009
期刊:
影响因子: --
作者:
M. Heindl;S. Tenbohlen;A. Kraetge;M. Krüger;J. Velásquez
通讯作者: J. Velásquez
DOI: --
发表时间: 2011
期刊:
影响因子: --
作者:
J. L. V. Contreras;M. A. Sanz;S. Banaszak;M. Koch
通讯作者: M. Koch
DOI: --
发表时间: 2007
期刊: IEEE PES PowerAfrica
影响因子: --
作者:
G. M. Kennedy;A. McGrail;J. Lapworth
通讯作者: J. Lapworth
清晰相似度和距离度量的实验分析
DOI: --
发表时间: 2014
期刊: International Conference of Soft Computing and Pattern Recognition
影响因子: --
作者:
Leila Baccour;R. John
通讯作者: R. John
频率响应分析(FRA)诊断变压器绕组异常的标准
DOI: --
发表时间: 2017
期刊:
影响因子: --
作者:
S. Miyazaki;Y. Mizutani;A. Taguchi;J. Murakami;N. Tsuji;M. Takashima;O. Kato
通讯作者: O. Kato