Research on Diagnosis Method of Transformer Winding Deformation Fault Based on Dynamic Time Warping

TAN Shan; ZHAO Zhongyong; YANG Jian; CAI Min; GUO Jinlong

doi:10.13296/j.1001-1609.hva.2024.02.011

您当前的位置：

首页 >

文章列表页 >

Research on Diagnosis Method of Transformer Winding Deformation Fault Based on Dynamic Time Warping

Research & Analysis | 更新时间：2025-04-26

- Research on Diagnosis Method of Transformer Winding Deformation Fault Based on Dynamic Time Warping
- High Voltage Apparatus Vol. 60, Issue 2, Pages: 108-118(2024)
- 作者机构：
  
  1.西南大学工程技术学院，重庆　 400715
  2.西南大学宜宾研究院，四川　宜宾　 644005
  3.国网新疆综合能源服务有限公司，乌鲁木齐　 830011
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(51807166);Fundamental Research Funds for the Central Universities(2023NSFSC0829)
- DOI：10.13296/j.1001-1609.hva.2024.02.011
  CLC：
- Received：09 September 2023，
  
  Revised：06 November 2023，
  
  Published：16 February 2024
- 稿件说明：
移动端阅览
TAN Shan, ZHAO Zhongyong, YANG Jian, et al. Research on Diagnosis Method of Transformer Winding Deformation Fault Based on Dynamic Time Warping[J]. High voltage apparatus, 2024, 60(2): 108-118.
DOI：

TAN Shan, ZHAO Zhongyong, YANG Jian, et al. Research on Diagnosis Method of Transformer Winding Deformation Fault Based on Dynamic Time Warping[J]. High voltage apparatus, 2024, 60(2): 108-118. DOI： 10.13296/j.1001-1609.hva.2024.02.011.

摘要

绕组变形是导致变压器故障的主要原因之一，频率响应分析法是一种常用的检测绕组变形故障的方法。文中针对实际应用中，因频率响应数据解释不足导致的绕组故障诊断效果不佳、抗噪性能差和故障程度指标与实际故障程度的单调性不良等问题，提出了基于动态时间

DTW

（dynamic time warping）规整路径与K最邻近算法（KNN，K-nearest neighbor）的变压器绕组状态判别法、基于

DTW

偏离度的变压器绕组故障程度表征法。通过在一台实际变压器及一台模型变压器上的运用，验证了其在绕组状态判别及绕组故障程度表征方面的性能。通过对比实验，分析了该方法在绕组状态判别中的准确性，抗噪性，以及在故障程度表征上的灵敏性与线性相关性。结果表明，在这两个案例中，与现行变压器绕组故障诊断标准相比，文中方法有更高的准确率，更能反映变压器绕组变形故障的程度，有着更好的抗噪性能。

Abstract

Winding deformation is one of a primary causes of transformer failures. Frequency response analysis (FRA) is a common method for detecting such deformation fault. In view of such issues as weak fault diagnosis effect

weak noise resistance and imperfect monotonicity fault degree index and and actual fault degree due to insufficient explanation of frequenc

y response data in the actual application

the condition discrimination method of transformer based on dynamic time warping (

DTW

) and the K-nearest neighbor (KNN) algorithm as well as fault characterization method leveraging

DTW

deviations are proposed. Based on the operation on a set of actual transformer and a set of model transformer

their performance in the aspect of winding status identification and characterization of fault degree of the winding are verified. The accuracy

noise resistance of the method in the winding condition identification and sensitivity as well as linear correlation in the characterization of fault degree are analyzed through comparison and experiment. The result shows that in these two cases the method

compared with the existing winding fault diagnosis standard of transformer

has higher accuracy

can reflect the degree of winding deformation of transformer and has better noise resistance performance.

关键词

Keywords

references

Views

179

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Influencing Factors on Deformation Diagnosis of Transformer Windings

Application of the Complex Values of the Frequency Response in Fault Diagnosis of Power Transformer

Transformer Winding Deformation Classification Method Based on Particle Swarm Algorithm Optimizing Support Vector Machine

Fault Diagnosis Model of Transformer Based on GJO Feature Optimization and AO-RF

New Approach to Transformer Fault Diagnosis Based on ANN

Related Author

CAO Xiao-long1

HU Chun-mei2

CAO Xiao-hu2

ZHANG Zhi-wu2

CHEN Ping1

LIU Dong-liang1

ZHOU Bin3

LIU Yunpeng

Related Institution

Gansu Province Tianshui Power Supply Company

Gansu Province Power Grid Maintenance Overhaul Company

Gansu Province Longnan Power Supply Company

Hebei Provincial Key Laboratory of Power Transmission Equipment Security Defense, North China Electric Power University

State Grid Shaanxi Electric Power Company

⁰