基于振动分析法的变压器故障分类和识别.docVIP

基于振动分析法的变压器故障分类和识别* 夏玉剑1，李敏2，陈果2，石同春2，沈大千2，王昕1 上海交通大学 电工与电子技术中心，上海 200240；2. 国网四川省电力有限公司广安供电公司，四川 广安 638500） 摘要：为了实现变压器故障的直观分类和故障识别，在分析变压器振动机理的基础上，本文提出一种基于主成分分析和KNN分类识别的变压器故障测方法。该方法采用EMMD(集合经验模式分解)方法提取变压器不同运行状态下振动信号的特征矢量，将该特征矢量通过主成分分析投影到直观的二维图像中。利用KNN分类识别实现故障分类和自动故障识别。试验结果表明，该方法可以实现对变压器正常状态、绕组变形、铁芯故障3种状态直观分类，并对测试样本进行快速的自动模式识别。 关键词：振动分析法；集合经验模式分解；特征矢量；主成分分析；K近邻法 中图分类号：TM403.9 文献标识码：A 文章编号：1001-1390（2016）00-0000-00 The classification and recognition of transformer fault based on vibration analysis Xia Yujian1, Li Min2, Chen Guo2, Shi Tongchun2, Shen Daqian2, Wang Xin1 (1.Center of Electrical and Electronic Technology, Shanghai Jiao Tong University, Shanghai 200240, China. 2. Guangan Power Supply Company of Sichuan Electric Power Co., Ltd., State Grid Corporation of China, Guangan 638500, Sichuan, China) Abstract: In order to achieve the fault identification and classification intuitively of transformer fault, this paper proposes a method of transformer fault detection method based on PCA (principal component analysis) and KNN (K-Nearest Neighbor) classification and recognition. In this paper, vibration signals from different transformer states are decomposed by EMMD (ensemble empirical mode decomposition) to abstract feature vectors which are projected onto a visual two-dimensional image. KNN classification is applied to verify fault classification and achieve automatic fault identification. Experimental results show that this method can achieve classification of a normal state of transformer, winding deformation and the core fault respectively, which can realize automatically pattern recognition of test sample. Keywords: vibration analysis, ensemble empirical mode decomposition, feature vector, principal component analysis, K-Nearest Neighbor (0 引 言 在电力系统的各种设备中，变压器是昂贵且重要的电力设备，其运行状态的可靠性是电网安全运行重要保障。统计表明，变压器发生故障较多的部件是其绕组和铁芯，此类故障事件在变压器总事故中所占比例达到70%[1]。因此，迫切需要一种可以准确判断变压器绕组和铁芯运行状态的方法。 目前，检测变压器运行状态的方法日益增多。文献[2-3]提出了短路阻抗法，在离线状态下，测量变压器绕组的阻抗值，分析其阻抗值变化