基于振动信号分形维数的变压器松动诊断方法

基于变压器松动前后振动信号非线性特性的改变,提出采用振动信号时域波形的分形维数作为铁心、绕组松动诊断特征量。文中探讨了分形维数应用于变压器振动信号分析的可行性,根据分形理论,针对变压器振动模拟信号与实测信号分别进行了分形盒维数计算。结果表明,正常状态下不同变压器绕组振动信号分形盒维数集中于1.19,不同变压器铁心振动信号分形盒维数不同,但对于同一台变压器,其铁心振动信号分形盒维数具有稳定值。变压器松动故障后,铁心和绕组振动信号的分形维数均明显增大,分形维数能准确反映变压器正常与松动状态。

Transformer fault diagnosis method based on fractal dimension of vibration signals

Based on the change of non-linear characteristics of vibration signals before and after the transformers looseness,a transformer fault diagnosis method based on the fractal dimension of the vibration signals time domain waveform was proposed to adopt as the diagnostic characteristic of iron core and winding looseness.This paper discussed the feasibility of the application of fractal dimension in the analysis of transformer vibration signals.In this paper,kinds of transformers vibration signals were measured and the fractal dimension of the vibration signals was calculated according to fractal theory.The results show that the fractal dimension of different kinds of winding signals of focus on 1.19 under normal state.The fractal dimension of iron core vibration varied with transformers,while it became stable with the same transformer.The fractal dimension of iron core and winding vibration signals significantly increased after the transformer looseness,it can accurately reflect the normal and loose state of transformer.