Diagnosis of Power Transformer Winding Deformation Based on Centroid Analysis of 3D Frequency Response Curve-Reference-Cited by-同舟云学术

Diagnosis of Power Transformer Winding Deformation Based on Centroid Analysis of 3D Frequency Response Curve

Published:2022-11 Issue:7 Volume:15 Page:567-578
ISSN:2352-0965
Container-title:Recent Advances in Electrical & Electronic Engineering (Formerly Recent Patents on Electrical & Electronic Engineering)
language:en
Short-container-title:EEENG

Author:

Tang Xuan¹^ORCID,Li Zhenhua¹

Affiliation:

1. College of Electrical Engineering and New Energy, China Three Gorges University, Yichang 443002, China

Abstract

Background: As the frequency of transformer winding faults becomes higher and higher, the frequency response analysis used to detect the winding status has attracted more and more attention. At present, there is still a lack of reliable and intelligent technologies for detecting the state of transformer windings in this field. Methods: In this paper, a transformer winding deformation diagnosis method based on centroid analysis of 3D frequency response curve is proposed. A 3D coordinate system is established with frequency, amplitude, and phase as the x, y, and z, respectively. The proposed method calculates the centroid of each 3D frequency response curve through an algorithm, and establishes a 3D coordinate system with the centroid of the normal winding 3D frequency response curve as the origin. The deviation distance and deviation angle between the centroid of mass of the 3D frequency response curve of the faulty winding in different frequency bands and the centroid of the 3D frequency response curve of the normal winding and the correlation coefficient of the traditional Frequency Response Analysis (FRA) curve. are calculated. The deviation distance and angle of the centroid and the correlation coefficient of the calculated frequency band are input as features into the Support Vector Machine (SVM) for intelligent diagnosis to achieve fault classification and optimizing its parameters with three optimization algorithms respectively. Results: The fault classification is realized by analyzing the distribution intervals of the centroids of the 3D frequency response curves of the three simulated faults. The accuracy of fault diagnosis using the 3D frequency response curve feature is significantly higher than the accuracy when using the traditional FRA curve feature and it proves the effectiveness of the proposed method. Conclusion: The proposed model can effectively identify different fault types and the diagnostic rate of the fault type is 100%.

Funder

National Natural Science Foundation of China

Educational Commission of Hubei Province of China

Publisher

Bentham Science Publishers Ltd.

Subject

Electrical and Electronic Engineering,Electronic, Optical and Magnetic Materials

Reference20 articles.

1. Weng H.; Chen H.; Wan Y.; A novel criterion to distinguish inrush current from fault current based on the Bhattacharyya coefficient. Power Sys Protect Control 2020,48(10),113-122

2. Krishnamurthy S.; Elenga Baningobera B.E.; “IEC61850 standard-based harmonic blocking scheme for power transformers”, Protect. Control Modern Power Sys 2019,4(2),121-135

3. He X.; Zhang Y.; Cui G.; Research on transformer fault detection method based on a regression algorithm. Power Sys Protect Control 2020,48(21),132-139

4. Fang T.; Qian Y.; Guo C.; Research on transformer fault diagnosis based on a beetle antennae search optimized support vector machine. Power Sys Protect Control 2020,48(20),90-96

5. Li Z.H.; Zhang Y.; Yao W.; Transformer winding micro deformation classification method based on sweep frequency impedance method and support vector machine In: Research Gate vol. 58, no. 1, pp. 99, 2019.