基于小波变换和概率神经网络的励磁涌流识别

针对变压器的励磁涌流问题,提出了一种基于小波变换和概率神经网络的新的变压器差动保护方案,用以实现励磁涌流与其它内部短路电流诸如单相短路、两相接地短路、三相接地短路、匝间短路的鉴别。利用小波变换进行信号分解,提取各尺度高频部分的能量,作为神经网络的输入特征向量,概率神经网络是为了进行模式识别。利用Matlab/Simulink平台上的仿真建模,获取励磁涌流和内部故障电流的数据。大量仿真结果显示,该方案可以有效地识别励磁涌流。

Identifying Transformer Inrush Current Based On Wavelet Transform and PNN

In this paper a new transformer differential protection scheme based on Wavelet Transform(WT) and Probabilistic Neural Network(PNN) is presented to identify inrush current from inner short-circuit current such as single-phase short-circuit current,two-phase grounding fault current,three-phase grounding fault current and turn-to-turn short circuit.WT is used for decomposition of signals,and the high frequency part energy of all scales is extracted as the neural network input feature vector.PNN for classification.Inrush current data and other transients are obtained by simulation using Matlab.Results show that the proposed procedure is efficient in identifying inrush current from other events.