基于最小二乘矩阵束的励磁涌流识别方案

准确、快速地切除变压器故障直接影响电力系统能否持续安全稳定运行,励磁涌流的鉴别正是变压器保护中的重要一环。为了有效辨识励磁涌流与故障电流,避免变压器差动保护的不正确动作,利用变压器励磁涌流与短路故障电流频域信息的差异,提出了一种基于最小矩阵束算法的励磁涌流识别新方案,以变压器两侧的差动电流作为信号量,通过矩阵束算法分析采样信号中不同频率的分量,计算含衰减因数的电流能量信息熵识别励磁涌流。仿真试验证明了新判据具有原理清晰、抗干扰能力强等优点,能够正确区分励磁涌流和故障电流,为后续研究及工程应用提供借鉴和参考。

Identification method for inrush current based on least-square matrix pencil algorithm

Accurate and rapid removal of transformer faults is related to the safe and stable operation of power system. The identification of magnetizing inrush current is a vital problem in transformer protection. In order to detect internal fault current from inrush current correctly and prevent incorrect operations of relays. By analyzing and summarizing the difference characteristics in frequency domain between fault current and inrush current, this paper proposed a novel inrush currents identification criterion based on least-square matrix pencil, using transformer differential current as detection signal. The matrix pencil algorithm is used to calculate the harmonic component of transformer current. And then entropy of current content and decay factor is computed to discriminate the inrush current. Simulation results show this scheme reliability and effective, can precisely identify the inrush current from the inner fault current. On this basis, key issues in the field are provided for the further research and engineering applications.