Fault Diagnosis of OLTC Based on Time-Frequency Image Analysis of Vibration Signal

Abstract

On-Ioad tap changer (OLTC), as a frequently operated part of transformer, is affected by both transient mechanical stress and electric stress, bringing high failure rates. The vibration signal is closely related to the mechanical state of the switch contact. Therefore, it is of great significance to monitor the contact state by analyzing the vibration signal. In this paper, the accelerometer is used to collect vibration signals during the switching operation of OLTC. The envelope of vibration signal is extracted by Hilbert transform. The region of sudden change is distinguished by singular value for each peak of envelope signal. The vibration signals generated by the action of the drive motor and switching process are divided into 5 similar parts according to the operation mode of OLTC. The time-frequency analysis images are obtained by using wavelet transform. The energy-mean of each part of the time-frequency analysis images are obtained. Comparing the energy-mean, it is found that the energy-mean of the 5 parts is similar. The results show that the mean-energy of the time frequency analysis image of the vibration signals of the drive motor and switching process can be used as the characteristic quantity, reflecting one the different changes of the vibration signals. Therefore, the mean-energy of the time-frequency analysis images of the vibration signals can be used for the mechanical state detection of the OLTC.