The Application Technique on AI and Statistical Analysis of 3d-PD

3d-PD의 통계적 고찰과 신경망 응용기술

  • Lim, Jang-Seob (Faculty of M.E.C., Mok-Po National Maritime University) ;
  • Park, Yong-Sik (Faculty of M.E.C., Mok-Po National Maritime University) ;
  • Choi, Byoung-Ha (Faculty of M.E.C., Mok-Po National Maritime University) ;
  • Han, Sok-Kyun (Faculty of M.E.C., Mok-Po National Maritime University)
  • 임장섭 (목포해양대학교 전자통신공학부) ;
  • 박용식 (목포해양대학교 전자통신공학부) ;
  • 최병하 (목포해양대학교 전자통신공학부) ;
  • 한석균 (목포해양대학교 전자통신공학부)
  • Published : 2001.05.11

Abstract

The partial discharge testing is widely used in diagnostic measuring technology because it gives low stress to power equipment which is undertaken tests. Therefore it is very useful method compare to previous destructive methods and effective diagnosis method in power system that requires on-line/on-site diagnosis. But partial discharges have very complex characteristics of discharge pattern, so it is required continuous research to development of precise analysis method. In recent, the study of partial discharge is carrying out discover of initial defect of power equipment through condition diagnosis and system development of degradation diagnosis using HFPD(High Frequency Partial Discharge) detection. In this study, simulated system is manufactured and HFPD occurred from those simulator is measured with broad-band antenna in real time, the degradation grade of system is analyzed through produced patterns in simulated target according to the AI/statistics processing.

Keywords