Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4945713 | International Journal of Electrical Power & Energy Systems | 2017 | 8 Pages |
Abstract
Frequency response analysis (FRA) is an effective technique to detect mechanical faults inside a power transformer. The FRA interpretation is normally carried out by comparing an FRA trace corresponding to a specific connection scheme with a fingerprint of its type. This paper, by contrast, proposes a novel method based on comparing FRA traces of different connection schemes recorded from the same winding to detect the mechanical deformations. In this contribution, the theoretical background is explained, and the experimental results are addressed to support the functionality of the proposed algorithm. Furthermore, a circuit model of the windings is created, and the simulation results are also used to validate the experimental measurements. Finally, two case studies describe the application of the new method in the condition assessment of transformers to achieve a more reliable decision. Although more research is needed to support the new applications, the early potential of the method is shown in this contribution.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Mohammad Hamed Samimi, Amir Abbas Shayegani Akmal, Hossein Mohseni, Stefan Tenbohlen,