煤矿变压器励磁涌流识别新方法

New Method to Identify Transformer Inrush Current in Coal Mine

针对煤矿变压器励磁涌流导致差动保护误动的问题,综合考虑了变压器励磁涌流和故障电流的特征,提出了一种基于粒子群算法的BP神经网络的新算法。利用MATLAB仿真软件对一个双端供电系统进行仿真,将检测到的电流信号作为PSO-BP神经网络的输入样本,进行训练及测试。仿真结果表明,此方案可以快速地、可靠地识别变压器励磁涌流。

problem of differential transformer protection malfunction caused by inrush current in some coals, the paper considering the characteristics of transformer inrush current and fault current, a new method to identify inrush current based on PSO-BP neural network is proposed. The simulation model of a double-ended power supply system is established by MATLAB. The current signals were obtained by simulation results and regarded as the PSO-BP neural network input samples, then training and testing. The simulation result shows that this method can identification transformer magnetizing inrush current and quickly and reliably. Key words： BP neural network; particle swarm optimization; inrush current; transformer