부분방전 데이터 배경잡음 전처리 알고리듬 개발

This study has been conducted to develop the pre-processing algorithm which calculates the BN (Background Noise) level and automatically removes it from PD (Partial Discharge) data. In this study, PD data was transformed into ordinary histogram to analyze the frequency of PD data and compute its BN level. And the algorithm was implemented by using the calculated BN level. The developed algorithm was applied to PD pattern recognition for the field PD data with BN. The results show that the pre-processing algorithm can effectively remove the BN and improve the reliability of pattern recognition.