基于波形系数的变压器励磁涌流快速识别算法

通过分析励磁涌流波形与磁化曲线的关系,提出了一种基于波形系数的变压器励磁涌流快速识别算法。利用空载合闸和内部故障后符合正弦特征的采样数据预测出参考波形,并将1／2周波内的采样数据与预测数据进行比较,以此定义波形系数。通过波形系数的大小区分变压器励磁涌流状态和内部故障状态。仿真和动模实验表明,该算法能在一个周波内准确识别出空载合闸和各种内部故障状态。

A Fast Algorithm to Identify Inrush Current Based on Waveform Factor

By means of analyzing the relationship of inrush current waveform and magnetization curve, a waveform factor based fast algorithm to distinguish inrush current of power transformer from its fault currents is proposed. In this algorithm the reference waveform can be forecasted by use of the data, which conform to sinusoidal characteristic and are sampled after no-load switching and internal faults respectively; and the sampled data within one half period are compared with forecasted data; then the waveform factor is defined by analyzing the similarities between practical sampled waveform and the forecasted standard sinusoidal waveform. According to the value of waveform factor the inrush current of transformer can be distinguished from its internal faults. The results of both digital simulation and dynamic simulation show that the proposed algorithm can correctly identify no-load switching and internal faults in the duration of one period.