Abstract
Due to its high sensitivity to small amounts of electrical faults, dissolved gas analysis (DGA) is a popular method for identifying faults in power transformers. The early prediction of fault type in transformers is a crucial part of power system reliability. As a result of this, it is extremely essential to keep a close eye on the behaviour of the power transformer while in service. Periodic monitoring of the condition of power transformers is important to prevent power outages. Early fault detection can result in substantial operational and maintenance cost savings, as well as the prevention of any early breakdown or failure. The current research study aims to assist experts in precisely diagnosing power transformer faults by providing an alternative to the drawbacks that classical DGA methods viz. Dornensburg ratio, Rogers' ratio, IEC ratio, and Key gas possess i.e., the lack of a proper diagnosis for instances that do not fall under the usual DGA codes. This work proposes an artificial intelligence (AI)-based fault diagnoses algorithm, which is a Support Vector Machine (SVM) for improving the diagnostic accuracy of existing DGA methods. According to the author's knowledge, there are no existing research works that have developed an SVM-based fault classification approach using Google Colab. As a result, the proposed work contributes to the fault diagnostics of power transformers. The results presented in this research showed that the proposed method SVM has high accuracy in diagnosing faults in power transformers as compared to the four methods presented.
Original language | English |
---|---|
DOIs | |
Publication status | Published - 2023 |
Event | 31st Southern African Universities Power Engineering Conference, SAUPEC 2023 - Johannesburg, South Africa Duration: 24 Jan 2023 → 26 Jan 2023 |
Conference
Conference | 31st Southern African Universities Power Engineering Conference, SAUPEC 2023 |
---|---|
Country/Territory | South Africa |
City | Johannesburg |
Period | 24/01/23 → 26/01/23 |
Keywords
- DGA
- faults
- methods
- power transformers
- SVM
ASJC Scopus subject areas
- Computer Networks and Communications
- Energy Engineering and Power Technology
- Renewable Energy, Sustainability and the Environment
- Electrical and Electronic Engineering
- Safety, Risk, Reliability and Quality