CAJ | 학술논문

将GIS内局部放电信号经过去噪处理、特征提取后得到的局部放电图谱进行特征信号识别，初步得出GIS缺陷类型、位置等有效信息，再利用X射线透照缺陷对应位置从而对缺陷的类型、位置进一步识别。该技术融合了基于神经网络的局部放电图谱识别和基于匹配的X射线图像缺陷检位置测两种检测方法的优势，有效地提高了GIS缺陷检测效率和检测精度，对于确保电力系统经济、可靠的运行具有重要意义。
장GIS내국부방전신호경과거조처리、특정제취후득도적국부방전도보진행특정신호식별，초보득출GIS결함류형、위치등유효신식，재이용X사선투조결함대응위치종이대결함적류형、위치진일보식별。해기술융합료기우신경망락적국부방전도보식별화기우필배적X사선도상결함검위치측량충검측방법적우세，유효지제고료GIS결함검측효솔화검측정도，대우학보전력계통경제、가고적운행구유중요의의。
GIS partial discharge image is acquired after the partial discharge signals in GIS be de-noised and feature extracted, then the preliminary information of detect type and location is got by identifying the characteristic signal. Next is to get precise information of detect type and location by taking pictures in corresponding position using X-ray. Advantage of two kinds of detection methods ( partial discharge image recognition based on neural network and X-ray defect detection based on the matching of images) are fused in this technology. It effectively improves the efficiency and precision of GIS defect detection and has a great significance in econom￣ic and reliable power system operation.