基于综合小波奇异距离的谐振接地系统故障选线

Fault line identification based on comprehensive wavelet singular distance in resonant grounded system

针对现有谐振接地系统故障选线准确率低等问题,提出一种基于综合小波奇异距离的自适应权重故障选线方法。将各线路零序故障电流进行归一化处理,分别提取其各尺度下的平稳小波奇异信息矢量,计算同一尺度下各出线平稳小波奇异信息矢量间的标准欧式距离相对系数矩阵,作为相应频段各出线故障奇异信息差异性的度量,通过自适应权重综合测度函数实现多尺度故障奇异信息的融合,从而选出故障线路。在不同过渡电阻、故障位置、噪声干扰和运行工况等情况下分别对所提方法进行仿真验证,结果表明,所提出的算法选线可靠、准确,适应性强。

Aiming at the problems such as low accuracy rate of fault line identification in the existing resonant grounded systems,a self-adaptive weight fault line identification method based on comprehensive wavelet singular distance is proposed.First,the zero-sequence fault current of each line is normalized before singular value decomposition.Then,the vectors of stationary wavelet singular information at each scale are extracted,and the standard Euclidean distance relative coefficient matrix between stationary wavelet singular information vectors of each line at the same scale is calculated,as a measure of fault singular information differences for feeders in respective frequency band.Finally,the fault feeder is identified through self-adaptive weighted comprehensive measure function calculation to realize the fusion of multi-scale fault singular information.Under the conditions of different transition resistances,different fault locations,noise interferences and different operating states,the simulation and verification of the proposed method are carried out respectively.The results show that the proposed algorithm is reliable,accurate and highly adaptable in fault line identification.