Yongwen Li , Bin Song , Shaocheng Wu , Deyu Nie , Zisheng Zeng , Linong Wang
{"title":"An innovative transformer fault diagnosis method based on ACK-GRANDE","authors":"Yongwen Li , Bin Song , Shaocheng Wu , Deyu Nie , Zisheng Zeng , Linong Wang","doi":"10.1016/j.epsr.2025.112239","DOIUrl":null,"url":null,"abstract":"<div><div>To address the limitations of existing DGA-based transformer fault diagnosis methods and further enhance fault diagnosis accuracy, this paper proposed a new transformer fault diagnosis method. First, during the pre-processing of DGA data, data cleaning was performed to remove anomalies and noise, and the Adaptive Synthetic Sampling Approach for Imbalanced Learning (ADASYN) algorithm was used to alleviate data imbalance, thereby improving data quality for subsequent diagnostics. Next, an improved model, referred to as ACK-GRANDE was proposed. The key optimization modules were divided into the following parts. Firstly, a refined feature weight allocation strategy was introduced, simulating a graph attention mechanism to adjust feature weights, avoid local optima, and enhance the ability to capture key information. Additionally, supplementary losses based on cosine similarity and Euclidean distance from K-Means clustering were incorporated, constructing a new loss function to enhance the model’s ability to recognize patterns in complex data. Finally, using 1200 DGA data samples with known fault types for case analysis, the results demonstrated that, compared to other traditional machine learning classification models, the proposed algorithm achieved higher accuracy in transformer fault diagnosis. The overall accuracy reached 93.75 %, exhibited superior long-term stability and broader applicability, and ablation experiments further validated the feasibility of each optimization module.</div></div>","PeriodicalId":50547,"journal":{"name":"Electric Power Systems Research","volume":"251 ","pages":"Article 112239"},"PeriodicalIF":4.2000,"publicationDate":"2025-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Electric Power Systems Research","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0378779625008260","RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
To address the limitations of existing DGA-based transformer fault diagnosis methods and further enhance fault diagnosis accuracy, this paper proposed a new transformer fault diagnosis method. First, during the pre-processing of DGA data, data cleaning was performed to remove anomalies and noise, and the Adaptive Synthetic Sampling Approach for Imbalanced Learning (ADASYN) algorithm was used to alleviate data imbalance, thereby improving data quality for subsequent diagnostics. Next, an improved model, referred to as ACK-GRANDE was proposed. The key optimization modules were divided into the following parts. Firstly, a refined feature weight allocation strategy was introduced, simulating a graph attention mechanism to adjust feature weights, avoid local optima, and enhance the ability to capture key information. Additionally, supplementary losses based on cosine similarity and Euclidean distance from K-Means clustering were incorporated, constructing a new loss function to enhance the model’s ability to recognize patterns in complex data. Finally, using 1200 DGA data samples with known fault types for case analysis, the results demonstrated that, compared to other traditional machine learning classification models, the proposed algorithm achieved higher accuracy in transformer fault diagnosis. The overall accuracy reached 93.75 %, exhibited superior long-term stability and broader applicability, and ablation experiments further validated the feasibility of each optimization module.
期刊介绍:
Electric Power Systems Research is an international medium for the publication of original papers concerned with the generation, transmission, distribution and utilization of electrical energy. The journal aims at presenting important results of work in this field, whether in the form of applied research, development of new procedures or components, orginal application of existing knowledge or new designapproaches. The scope of Electric Power Systems Research is broad, encompassing all aspects of electric power systems. The following list of topics is not intended to be exhaustive, but rather to indicate topics that fall within the journal purview.
• Generation techniques ranging from advances in conventional electromechanical methods, through nuclear power generation, to renewable energy generation.
• Transmission, spanning the broad area from UHV (ac and dc) to network operation and protection, line routing and design.
• Substation work: equipment design, protection and control systems.
• Distribution techniques, equipment development, and smart grids.
• The utilization area from energy efficiency to distributed load levelling techniques.
• Systems studies including control techniques, planning, optimization methods, stability, security assessment and insulation coordination.