基于QBVMD的变压器励磁涌流与故障识别研究

Research on Transformer Inrush Current and Fault Identification Based on QBVMD

摘要: 针对变压器差动保护易受涌流干扰而误动问题，本文提出了基于变分模态分解与优化神经网络的励磁涌流辩识方法。运用PSCAD软件平台搭建变压器励磁涌流与故障电流模型，获取涌流数据，利用准二元变分分解QBVMD对差流信号进行分解，获得不同尺度的信号后，在利用优化神经网络对故障类型进行辨识。最后，设置变压器励磁涌流、和应涌流、单相接地和三相接地故障情景并获得相应的故障数据，利用上述方法对故障类型进行辨别，结果表明，论文方法可以准确的识别变压器励磁涌流和故障电流，识别准确率高，为辩识变压器涌流与故障提供了参考。

Abstract: In order to solve the problem that transformer differential protection is easy to be disturbed by inrush current, this paper proposes an identification method of inrush current based on variational mode decomposition and optimized neural network. Using PSCAD software platform to build transformer inrush current and fault current model, obtain inrush current data, usequasi-bi- VMD to decompose differential current signal, obtain different scale signal, then use optimized neural network to identify fault type. Finally, the transformer magnetizing inrush current, reactive inrush current, single-phase grounding and three-phase grounding fault scenarios are set up and the corresponding fault data are obtained. The results show that the proposed method can accurately identify transformer inrush current and fault current, and the recognition accuracy is high, which provides a reference for identifying transformer inrush current and fault.