Mechanical Condition Monitoring of On-load Tap Changer by Tensor Decomposition with Vibration Signals
- Resource Type
- Conference
- Authors
- Cheng, Deliang; Wang, Fenghua
- Source
- 2021 IEEE Sustainable Power and Energy Conference (iSPEC) Sustainable Power and Energy Conference (iSPEC), 2021 IEEE. :3867-3872 Dec, 2021
- Subject
- Power, Energy and Industry Applications
Vibrations
Employee welfare
Tensors
Phase measurement
Vibration measurement
Transformer cores
Feature extraction
on-load tap-changer
feature extraction
phase space reconstruction
tensor decomposition
- Language
Vibration signals have becoming an important indicator to assess the mechanical condition of transformer on-load tap-changer (OLTC). Based on the chaotic features of vibration signals resulted from more than one measuring channels, the phase space three-order tensor is built in this paper according to the results of phase space reconstruction. Then the Tucker tensor decomposition is applied to extract the core tensor to identify the different mechanical conditions of OLTC. The vibration signals of a CM type OLTC model under normal and different fault conditions are measured. The calculated results have shown that the core tensor obtained by Tucker decomposition is capable of accurately describing the mechanical condition information of OLTC. The results could provide an important reference for the mechanical condition diagnosis of OLTC by vibration analysis methods.