Scholarlink

<p>Traditional Dissolved Gas Analysis (DGA) interpretation methods such as Roger’s ratios and International Electrotechnical Commission (IEC) standards often classify many samples as “Not-Identified” due to their reliance on strict threshold values. These uncertainties limit their ability to accurately detect early-stage transformer faults. This study aims to develop a Fuzzy Logic (FL) diagnostic model that overcomes the “Not-Identified” problem in conventional methods, thereby improving fault classification and enabling earlier detection and maintenance actions in oil-immersed power transformers. A Mamdani-type FL system was applied to Roger’s and IEC interpretation methods to accommodate uncertainty in making crisp decisions. Sixty samples from different power transformers of the Egyptian Electricity Transmission Company (EETC) were examined for DGA analysis. The proposed fuzzy approach was used for the re-evaluation of cases that were previously determined as “Not-Identified” using the conventional method. The fuzzy model identified all of the samples (100%) by overcoming the “Not-Identified” cases and providing a definite fault type to all previously uncertain samples. In addition, it created a Fuzzy Severity Index (FSI) that can measure fault severity and helps maintenance schedule activities in order of risk. The developed fuzzy-mediated diagnostic system enhances the reliability and precision of transformer fault diagnosis, thus providing a practical and reliable tool for on-time maintenance planning and online condition monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Fuzzy logic–based transformer fault diagnosis employing roger’s and IEC techniques

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月24日

查看全文
  • Amir G. Abd El-Rahim,
  • Khaled H. Ibrahim,
  • Mokhtar S. Ibrahim,
  • Eslam M. Abd Elaziz

摘要

Traditional Dissolved Gas Analysis (DGA) interpretation methods such as Roger’s ratios and International Electrotechnical Commission (IEC) standards often classify many samples as “Not-Identified” due to their reliance on strict threshold values. These uncertainties limit their ability to accurately detect early-stage transformer faults. This study aims to develop a Fuzzy Logic (FL) diagnostic model that overcomes the “Not-Identified” problem in conventional methods, thereby improving fault classification and enabling earlier detection and maintenance actions in oil-immersed power transformers. A Mamdani-type FL system was applied to Roger’s and IEC interpretation methods to accommodate uncertainty in making crisp decisions. Sixty samples from different power transformers of the Egyptian Electricity Transmission Company (EETC) were examined for DGA analysis. The proposed fuzzy approach was used for the re-evaluation of cases that were previously determined as “Not-Identified” using the conventional method. The fuzzy model identified all of the samples (100%) by overcoming the “Not-Identified” cases and providing a definite fault type to all previously uncertain samples. In addition, it created a Fuzzy Severity Index (FSI) that can measure fault severity and helps maintenance schedule activities in order of risk. The developed fuzzy-mediated diagnostic system enhances the reliability and precision of transformer fault diagnosis, thus providing a practical and reliable tool for on-time maintenance planning and online condition monitoring.