부분방전 패턴인식에 대한 BP및 SOM 알고리즘 비교 분석

Abstract

SOM(SeLf Organizing Map) algorithm which has some advantages such as data accumulation ability and the degradation trend trace ability was compared with conventionally used BP(Back Propagation) algorithm. For the purpose, partial discharge data were acquired and analysed from the artificial defects in CIS. As a result, basically the pattern recognition rate of BP aJgorithm was found out to be better than that of SOM algorithm. However, SOM algorithm showed a great on-site-applicability such as ability of suggesting new-pattem-possibility. Therefore, through increasing pattern recognition rate it is possible to apply SOM algorithm to partial discharge analysis. Also, for the image processing method it is required the normalization of the PRPDA graph. However, due to the normalization both BP and SOM algorithm have shown worse results, so that it is required further study to solve the problem.