Analysis Of Transformer Surface Vibration Signal In Operation

Power system plays an important role in the national economic development.Transformer is an important equipment in power system.The operation state of transformer has a great influence on the safe and stable operation of power system.Current transformer on-line monitoring methods exist some defects and the insufficiency,on-line monitoring method based on vibration signal analysis can be under the condition of the normal operation of the transformer to analyze its actual running status of equipment easy and convenient to carry,easy to operate,cooperate with the traditional method,can effectively improve the safety and economy of the equipment operation.In this paper,time-frequency analysis of the measured signal on the transformer surface was carried out.The continuous wavelet scale spectrum and time-wavelet methods were used to analyze the signal in frequency domain.Cross wavelet scale spectrum was used to analyze the correlation of different measuring points in time-frequency domain,and it was concluded that the farther the distance between measuring points was,the lower the correlation was,and the transverse comparison between different measuring points was of little significance.For the data collected by the 3d sensor,in addition to the characteristic analysis in the frequency domain,this paper also analyzed the phase correlation between the signals in different directions based on the cross wavelet transform method,found that there was a certain regularity in the phase of the harmonic components of the fundamental frequency.It provided a new way to analyze transformer by vibration method.Finally,a method combining wavelet transform with convolutional neural network was proposed.The time-frequency graph after the wavelet transform was taken as the input part of the convolutional network,and the advantage of the convolutional neural network was utilized to classify and recognize it.This method is effective in the classification and recognition of transformer surface vibration signals and has a good application in the analysis of vibration signals.