基于波形时域分布特征的变压器励磁涌流识别

摘要

A new method of transformer inrush currents identification based on the distribution characteristics of current waveforms in time domain is presented in this paper.The current waveform distribution coefficient is defined based on the histogram,and the distribution characteristics of the differential currents under the conditions of inrush and internal fault are analyzed.Based on the above analysis the waveform distribution characteristic coefficient is defined as the ratio of the waveform distribution coefficients of two different subintervals,and whether the value of the coefficient is within certain region is used as the criterion to identify inrush currents.The theory basis of determining the number of subintervals is analyzed.The proposed method is verified with the data obtained from dynamic analog and digital simulation experiments.The results show that the method is able to discriminate between inrush currents and fault currents rapidly with less calculation,and the performance of the method in the aspects of reliability and sensitivity is superior to those of the second harmonic restraint principle and the dead angle principle.%提出了一种基于波形时域分布特征的新型变压器励磁涌流识别方法。文中在直方图技术的基础上定义电流波形分布系数,分析了变压器内部故障电流和励磁涌流的波形分布特征,在此基础上将两个不同子区间的分布系数的比值定义为波形分布特征系数,提出了根据该系数大小来识别励磁涌流的具体实现方案,并分析了确定子区间数目的理论依据。利用动模实验和数字仿真数据对本方法做了验证分析,结果表明：本方法计算量小,能快速区分励磁涌流和故障电流,在可靠性和灵敏性等方面的性能优于目前的二次谐波判据和间断角判据。