Research On Fault Diagnosis And Protection Of Power Transformer Based On Wavelet Transform

In the application of power transformer,differential protection is usually used as the critical protection,and the lower rate of right operation of the transformer differential protection is influenced by the appearance of the inrush current.Therefore,the fast and right identification is increasingly importance for inrush current,and it has always been a hot issue in the research of transformer protection.This paper makes a further research on the hot problems in the protection of power transformers.Meanwhile,this paper describes the method of differential protection firstly,and it also introduces the excitation inrush mechanism secondly.Finally,it summarizes identification methods usually used of the current inrush in recent years.Furthermore,this paper builds model of the inrush current and fault current simulation using power system toolbox in MATLAB to simulate various transformer states for data analysis.The simulation results show that the simulation results can correctly respond to all kinds of faults and magnetizing inrush state of transformer,and results provide data support for further research.Wavelet has strong ability of time-frequency and domain-frequency.First,this paper introduces the basic theory and application of wavelet transform.Second,a new judgment algorithm based on wavelet energy ratio is proposed in this paper.This paper uses db5 wavelet 5 scale wavelet transform to handle the simulation data.The wavelet analysis showed that there exist obvious differences in the energy fluctuation of inrush current and fault current in the wavelet transform in the frequency and the variation trend.Then this paper calculates the mean energy ratio K of each phase of a multi frequency harmonic scale under the new criterion based on energy ratio.Distinguishing inrush current and fault current by the size of K value.Meanwhile,the optimal threshold is selected through a lot of simulations.When the K value of the maximum phase is greater than the threshold,the inrush current is recognized.When the maximum phase's K value is less than the threshold,the fault current is recognized.Finally,a large number of experiments show that the criterion can in one cycle to accurately distinguish inrush current and fault current,and the performance is betterthan the traditional two harmonic braking method and previous energy criterion.And it has a high stability and right prospect in the application of power transformer.Finally,a differential protection system based on wavelet transform is designed for power transformer in this paper,and the system design based on STM32 platform which includes high computing capability,low power consumption,peripheral rich,digital signal processing capability of high real-time.This paper mainly introduces the system hardware overall structure diagram,acquisition module,processing and control output module and the display module.Meanwhile,it introduces the major module software structure design through the program flow chart.At the same time,it tests data processing algorithm on STM32F7 platform.The protection system has a practical value.