A New Kind of Transformer Oil State Detection Method based on Multi-frequency Detection Technique and Multivariate Statistics

DOI

10.2991/iccte-16.2016.206How to use a DOI?

Keywords

multi-frequency ultrasound; transformer oil; fault detection; multivariate statistical analysis; complex artificial neural network; self-learning

Abstract

The operating status of transformers directly affects the safety of entire power grid. Therefore, it is essential to monitor a variety of transformer fault and ensure the normal operation of power grid. The multi-frequency ultrasonic technique was proposed in the thesis and which utilized hundreds of different frequency ultrasonic signals to test transformer oil for transformer fault. Various ultrasonic parameters are received and based on which the multivariate statistical analysis method and the plural artificial neural network data analysis method are applied to obtain the characteristic values, and the operation state values are compared with the standard values stored in the background database, so that the evaluations can be made on transformer’s current operation situations. The design principle of the transformer oil measurement system was elaborated and a multivariate statistics method was given in the paper and which was applied to build a certain internal relationship between the ultrasonic parameters and stored sample values in database, and then the state of transformer that contained in the ultrasonic parameters can be released. An experimental test verifies that the developed system can effectively and accurately detect multiple transformer faults via detailed analysis of ultrasonic parameters of transformer oil, thus provides significant guidance and reference for the transformer detection departments.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Rui Rao
AU  - Zhengjia Li
AU  - Haoyong Song
AU  - Yuqing Chen
AU  - Dan Li
PY  - 2016/01
DA  - 2016/01
TI  - A New Kind of Transformer Oil State Detection Method based on Multi-frequency Detection Technique and Multivariate Statistics
BT  - Proceedings of the 2016 International Conference on Civil, Transportation and Environment
PB  - Atlantis Press
SP  - 1173
EP  - 1183
SN  - 2352-5401
UR  - https://doi.org/10.2991/iccte-16.2016.206
DO  - 10.2991/iccte-16.2016.206
ID  - Rao2016/01
ER  -