Summary
In the current global energy context, electrical power grids play a key role in the energy transition, by ensuring a stable and uninterruptible power supply. Among the most critical assets in the power grid, power transformers play an important role, among other things, in regulating voltage levels. This is possible through the On-Load Tap Changer (OLTC), a complex mechanical component that allows for voltage regulation under load, without interruption of power flow. Due to its mechanical nature and intensive use, it is the most failure-prone component, making OLTC failures one of the leading causes of transformer stoppages, often causing significant grid operational disruptions and financial losses. Traditional maintenance techniques typically fail to detect incipient faults. Considering this, predictive maintenance strategies sustained by continuous online condition monitoring have become increasingly valuable. This work introduces an advanced OLTC condition monitoring approach based on electrical signature analysis, enhanced with artificial intelligence, that makes use of a Siamese
Read more Read lessNeural Network (SNN) model where two parallel convolutional neural networks compare incoming impedance images with learned healthy and faulty reference patterns in order to compute a similarity score. Key parameters such as start time, end time, and duration of each tap transient, are combined with the similarity score computed before in order to have a global severity score index. This proposed condition assessment approach enables objective fault detection, contributing to a more resilient and intelligent power grid.
Additional informations
| Publication type | Session Materials |
|---|---|
| Reference | A2_11095_2026 |
| Publication year | |
| Publisher | CIGRE |
| Country | Portugal |
| Study committees | |
| File size | 700 KB |
| Price for non member | 30 € |
| Price for member | 30 € |
Authors
ABRANTES Gonçalo - ENGING; SANTOS Gonçalo - ENGING; DUARTE Rogério - ENGING; ESTIMA Jorge - ENGING; FERREIRA Marco - ENGING
