基于关联维数的牵引变压器励磁涌流鉴别研究  

Study on Identification of Magnetizing Inrush Current of Traction Transformer based on Correlation Dimension

作　　者：李久 周启斌 LI Jiu;ZHOU Qibin

机构地区：[1]中铁电气化局集团有限公司设计研究院

出　　处：《电气化铁道》2022年第5期15-19,共5页Electric Railway

摘　　要：由于牵引变压器差动保护在励磁涌流的冲击下可能误动作,精准识别励磁涌流和短路电流,可提高变压器继电保护的可靠性。本文研究通过建立高效的模态特征提取方法及关联维数算法,将牵引变压器差动电流信号经相空间重构和关联维数计算,从差动电流信号中准确捕捉到牵引变压器匝间的动态特征信息,并以该关联维数特征信息作为识别两种状态电流的依据。仿真实验结果证明,该方法能快速准确识别出变压器的内部短路故障,具有灵敏度高、可靠性强等优点,并且适用于复杂的牵引供电系统状态故障诊断。Since the misoperation may be occurred at differential protection of traction transformer under impacts of magnetizing inrush current,so the accurate identification of magnetizing inrush current and short circuit current can improve the reliability of transformer relay protection.On the basis of establishment of effective modal characteristics attraction method and correlation dimension algorithm,the phase-phase reconstruction and correlation dimension algorithm are made for the differential current signals of traction transformer,and the dynamic characteristic information between turns of traction transformer is accurately captured from the differential current signal,and the correlation dimension characteristic information is used as the basis for identifying the two states of the current.The simulation results verify that:the method is able to identify rapidly the inner short circuit faults of transformer,and it has advantages of high sensitivity,reliability and is applicable for the diagnosis of state and fault of complicated traction power supply system.

关 键 词：牵引变压器 关联维数 励磁涌流 特征矢量 故障诊断 

分 类 号：U226.7[交通运输工程—道路与铁道工程]