Characteristics Analysis And Extraction Of Surface Vibration Signals Of Running Transformer

Transformer is the key equipment of power system,and the reliability of its operation affects the security of power grid.Statistic shows that winding and core are the multiple components of transformer fault.Vibration signals on the surface of transformer are closely related to the running condition of winding and core.Therefore,through analyzing and extracting the characteristics of the surface vibration signals of transformer,the condition of the winding and core can be reflected accurately.And then the result could be used to monitor the operation state of the transformer by vibration method.Three-dimensional vibration signals on the surface of normal running transformer are studied in this paper.Combined with the load current and running voltage,the peak characteristics in time domain,the frequency spectrum of Fourier transform and the energy spectrum of wavelet packet in different directions are analyzed and summarized.On the basis of characteristic analysis and the demand of vibration signals feature extraction of transformer surface,the energy,correlation and sensitivity of the vibration signals of the transformer are analyzed combined with the data of the load current and running voltage.In this way,A feature vector which can be used to describe the variation of the operating state of winding and core is constructed.The influence of the change of load current and running voltage on the energy at each frequency can be quantified by the feature vector.This method provides a new idea for the feature extraction of transformer.Transformer operating conditions are complex and changeable,which have a significant impact on the surface vibration signals.Refer to the method of the wind turbine,the operation condition of transformer is divided by the interval division method.Utilizing the feature vector and energy distribution of the vibration signals under different working condition,the abnormal criterion of transformer is set,which can be used to monitor the state of transformer online.