中文摘要：

励磁涌流的畸变波形会呈尖顶波特征，而故障电流基本为基频正弦波，基于此提出一种利用改进数学形态学来识别变压器励磁涌流的新方法。该方法从形态开和闭运算的几何解释出发，采用数据延拓的方法克服开和闭运算中存在的“端点效应”问题，进而定义并构造出波峰谷最大算子函数MBT函数，针对正弦波形定量推导所选结构元素半径与MBT函数值之间的关系，构造指示波形凹凸特性的相对凹凸系数KCC并根据差流的凹凸系数KCC来区分变压器励磁涌流和内部故障。理论分析及动模实验结果表明该方法简单实用且判断速度快（判断时间是8-13ms），并具有较好的抗CT饱和性能，亦不受系统频率波动等的影响。

英文摘要：

A new algorithm mathematical morphology was based on presented improved for fast discrimination between inrush current and fault current of power transformers. The method is based on the principle that the waveform of the inrush distorts seriously, while the fault current keeps nearly sinusoid. By the geometric explanation of open operation and close operation of mathematical morphology, ＂endpoint effect＂ was overcome. On the basis of self-defined MBT （max-bottom-top-operator） function in mathe- matical morphology, this paper deduced the formula against sine wave which indicated the relationship between the selected structural elements radius and the value of MBT function, and constructed relative concavo-convex coefficient Kcc which was used for showing the concavo-convex characteristic. According to the Kco value of differential current, both inrush current and fault current of power transformers can be distinguished. Theoretical analysis and dynamic simulation results show that the method is effective under various fault conditions and simple to be applied. Its performance is hardly affected by CT saturation and frequency fluctuation. Moreover, judgement time of fault is 8-13 ms.