基于振动信号的变压器绕组松动故障诊断方法

发布时间：2018-07-14 11:01

【摘要】：提出了基于运行中变压器表面振动信号的绕组松动故障诊断模型和诊断方法。首先,分析了绕组振动幅值与电流、预紧力、铁芯振动、非线性因素的关系,确定振动信号中100 Hz为绕组松动的特征频率,提出分离绕组振动幅值和铁芯振动幅值的方法;提出绕组松动诊断模型和基于该模型的平均安全余量,利用待检测变压器在一组负载电流下的绕组振幅计算平均安全余量,由此定量判断绕组的松动状态。利用有限元仿真和现场实验分别对诊断模型和诊断方法进行检验,实验结果显示平均安全余量在1±0.5之外时存在松动故障,值越远离1,松动程度越大。该方法解决了变压器实际运行中绕组和铁芯振动基频相同、矢量叠加相互影响的问题及实际运行中负载变化,电流不同,仅用某个电流下的振幅判断精度受限的问题,且对松动程度量化表示,易于判断。
[Abstract]:The diagnosis model and diagnosis method of winding loosening fault based on the vibration signal of transformer surface in operation are presented. Firstly, the relationship between vibration amplitude of winding and current, pretension force, core vibration and nonlinear factor is analyzed. The characteristic frequency of winding loosening is determined in vibration signal 100Hz, and the method of separating vibration amplitude of winding and vibration amplitude of iron core is put forward. A diagnosis model of winding loosening and an average safety margin based on the model are proposed. The amplitude of the winding under a group of load currents is used to calculate the average safe margin, and the loosening state of the winding is quantitatively judged. The diagnosis model and diagnosis method are tested by finite element simulation and field experiment respectively. The experimental results show that there is loosening failure when the average safety margin is 1 卤0.5, and the loosening degree is greater as the value is far away from 1. This method solves the problem that the fundamental frequency of winding and core vibration is the same and the vector superposition affects each other in the actual operation, and the load changes in actual operation, the current is different, and the accuracy of judging the amplitude is limited only by the amplitude of a certain current. It is easy to judge the degree of looseness.
【作者单位】：河海大学能源与电气学院;国网江苏省电力公司南京供电公司;
【基金】：国网江苏省电力公司重点科技项目(J2015054)资助
【分类号】：TM407

【相似文献】