가스절연지상개폐기 PD 자동진단기법 연구 = A Study on the Automatic PD Diagnosis Method for Gas Insulated Load Break Switches

다국어 초록 (Multilingual Abstract)

Partial discharge diagnosis is being carried out to prevent failure of ground switch of distribution facilities. However, the reliability of partial discharge diagnosis is insufficient because the results are different depending on how the experts classify the partial discharge and noise signals. Therefore, in this paper, a method to automatically classify partial discharge and noise signals was studied in order to improve the reliability of the partial discharge diagnosis results of ground switch. Partial discharge and noise signal are obtained through the High Frequency Current Transformer(HFCT). And the features were extracted based on the pulse shape analysis, the standard deviation and gravity center of signal. The features are used as input to neural networks and learned using back-propagation. In addition, the structure of neural networks was optimized through genetic algorithm.

참고문헌 (Reference)

1 Honhseok Lee, "Trends of Development system for Power Facility Asset Management" 35 (35): 23-28, 2021

2 D. Benzerouk, "Pulse Sequence Analysis and Pulse Shape Analysis: Methods to Analyze Partial Discharge Processes"

3 J. J. Grefenstette, "Optimization of Control Parameters for Genetic Algorithms" 16 (16): 122-128, 1986

4 S. Chakraverty, "Artificial Neural Networks for Engineers and Scientist : Solving Ordinary Differential Equations" CRC Press 2017

5 Kang-Won Lee, "Artificial Neural Network Discrimination of Multi-PD Sources Detected by UHF Sensor" 3C (3C): 5-9, 2003