摘要
小电流接地系统发生接地故障后,若无法及时快速确定故障线路并加以切除,故障过电压长期作用有可能造成电力设备损坏,严重影响系统安全,而小电流接地系统接地故障信号不明显,为故障选线带来困难。鉴于此,针对小电流接地系统单相接地故障在选线时存在的信号微弱、特征不明显等问题,利用一种灰狼优化后的变分模态分解(VMD)提取故障选线特征,以提高信号的分解精度,并与神经网络进行联合实现故障线路的甄别。最后利用ATP-EMTP仿真实验验证所提方法的准确性,与小波算法及传统算法对比,优化后的系统选线准确率更高。
After small current grounding system encounters a grounding fault,if the fault line cannot be quickly determined and cut off in time,the long-term effect of fault overvoltage may cause damage to power equipment,seriously affecting system safety.Nevertheless the grounding fault signal of small current system is faint,which brings difficulties to fault line determination.In view of this,a grey-wolf-optimized variational modal decomposition(VMD)was introduced to extract the faint signal characteristics of single-phase grounding faults and improve accuracy of signal decomposition.And by combining the optimized VMD with neural networks,the identification of fault lines was realized.The proposed method was proved by ATP-EMTP simulation to have satisfactory accuracy higher than wavelet algorithm and traditional algorithms in determining fault line.
作者
何龙
朱咏明
王清彬
陆聪
HE Long;ZHU Yongming;WANG Qingbin;LU Cong(State Grid Xinjiang Electric Power Co.,Ltd.,Changji Power Supply Company,Changji 831100,China;Anhui Zhengguang Electric Power Technology Co.,Ltd.,Hefei 230000,China)
出处
《电工技术》
2024年第2期48-51,共4页
Electric Engineering
关键词
VMD
接地故障
选线算法
VMD
grounding fault
line determining algorithm
