基于VMD双狼群算法的电网单相接地故障选线优化

针对电网单相接地故障选线的问题,对传统小波分解线路故障选线法进行改进,提出一种VMD(Variational Mode Decomposition)双狼群算法对故障选线进行优化。利用变分模态分解VMD提高信号的分解精度,对各条线路故障前后的能量特征进行提取,并形成故障的特征向量作为神经网络的输入。同时为了提高传统狼群算法的寻优精度,引入双狼群算法,建立一种VMD-DLWCA-NN模型,并由该模型的输出来判定故障线路。仿真实验验证了该方法的准确性。通过与小波算法法及传统算法对比,表明优化后的系统选线效果更好。

SINGLE-PHASE GROUNDING FAULT LINE SELECTION OPTIMIZATION FOR POWER GRID BASED ON VMD DOUBLE WOLVES COLONY ALGORITHM

Aiming at the problem of single-phase grounding fault line selection in power network, we improved the traditional wavelet decomposition fault line selection method and proposed a VMD dual-wolf colony algorithm to optimize the fault line selection. The variational mode decomposition (VMD) was used to improve the signal decomposition accuracy. The energy features of each line before and after the fault were extracted and the fault eigenvector was formed as the input of the neural network. To improve the searching accuracy of traditional wolves algorithm, we introduced the double wolves colony algorithm, and established a VMD-DLWCA-NN model. The fault line was determined by the output of the model. The simulation experiment proves the accuracy of the method. Compared with wavelet algorithm and traditional algorithm, the optimized system is more effective in line selection.