基于小波理论和多分辨分析的变压器励磁涌流识别方法

提出一种基于小波变换和多分辨分析识别变压器励磁涌流和内部故障电流的新方法。根据故障初始阶段励磁涌流的高频分量随时间推移逐渐变大而故障电流高频分量逐渐变小的特点，首先用Daub-4小波对励磁涌流和故障电流进行分解，得到第1层小波系数的2个初始峰值点，然后通过引入可靠性系数来识别励磁涌流和故障电流。动模试验结果表明，该方法能够可靠快速地识别励磁涌流和故障电流，且不需要考虑保护的整定值，实现方便。

A Method to Identify Inrush Current of Transformer Based on Wavelet Theory and Multiresolution Analysis

Based on wavelet transform and multiresolution analysis, a new method to distinguish transformer inrush current from its internal fault current is proposed. According to the property that at the initial stage of fault the high frequency component of inrush current gradually increases with time and the high frequency component of fault current gradually decreases with time, in the proposed method firstly the inrush current and fault current are decomposed by means of Daub-4 wavelet; then two peak value points of the first layer of wavelet coefficients are obtained; and then by means of introducing reliability coefficient the inrush current and fault current can be identified. Results of dynamic simulation show that the proposed method can identify inrush current and fault current of power transformer quickly and reliably while the setting of protection devices is not to be considered, and the proposed method is easy to implement.