Parameter Estimation of Transformer Frequency Response Model Using Machine Learning Methods

• DOI: 10.5207/jieie.2022.36.2.008
• Publication Date: 2022-03-08T00:35:36Z
• Authors: Junhyun Im; Yeunggurl Yoon; Sungyun Choi
• Keywords: Dissolved Gas Analysis; Delta-wye transformer

Examining power transformer faults is crucial for maintaining the reliability of system. The most popular methods detecting fault include thermal analysis, vibration partial discharge dissolved gas analysis(DGA), and sweep frequency response analysis(SFRA). Especially, SFRA test examined to detect internal such as winding fault. Simulation-level analysis enables inspection before connecting grid. This paper proposes a parameter estimation method using machine learning equivalent model.