대한전기학회:학술대회논문집 (Proceedings of the KIEE Conference)
- 대한전기학회 2001년도 하계학술대회 논문집 A
- /
- Pages.467-469
- /
- 2001
초음파 전달특성과 신경망을 이용한 효과적인 부분방전 위치 추정
Effective Location Estimation of Partial Discharge Using Ultrasonic Signal's Propagation Characteristic and Neural Network
- Seo, In-Chul (Dept. of Electrical Eng., Soong-Sil University) ;
- Kim, Jae-Chul (Dept. of Electrical Eng., Soong-Sil University) ;
- Jeon, Young-Jae (Dept. of Electrical Eng., Soong-Sil University) ;
- Kim, Young-No (Dept. of Electrical Eng., Soong-Sil University)
- 발행 : 2001.07.18
초록
This paper presents an efficient location estimation methodology for a partial discharge(PD) source in the power transformer. The methodology previously proposed is not suitable for locating a PD source because the ultrasonic signal with medium attenuated through the inner structure. In general, the propagation characteristic of ultrasonic signal with medium is different. Therefore, using this propagation characteristic, we can detect the location of PD source. In this paper, the proposal algorithm finds the PD source using three sensors attached to one side of the transformer without attenuation of the ultrasonic signal and applying the neural network based geometrical method. The proposed methodology demonstrates the effectiveness and validity on an experimental transformer.
키워드