High Impedance Fault Detection Method Based on Variational Mode Decomposition and Teager-Kaiser Energy Operators for Distribution Network

作     者：Wang, Xiaowei Gao, Jie Wei, Xiangxiang Song, Guobing Wu, Lei Liu, Jingwei Zeng, Zhihui Kheshti, Mostafa 

作者机构：Xi An Jiao Tong Univ Sch Elect Engn Xian 710049 Peoples R China Henan Polytech Univ Sch Elect Engn & Automat Jiaozuo 454000 Henan Peoples R China Zhejiang Elect Power Co State Grid Wenzhou Power Supply Co Wenzhou 325000 Peoples R China Tech Univ Berlin Fac Elect Engn & Comp Sci D-10623 Berlin Germany Zhuhai Xuji Elect Co Dept Res & Dev Zhuhai 519000 Peoples R China Shandong Univ Key Lab Power Syst Intelligent Dispatch & Control Jinan 250061 Peoples R China 

出 版 物：《IEEE TRANSACTIONS ON SMART GRID》 (IEEE智能电网汇刊)

年 卷 期：2019年第10卷第6期

页      面：6041-6054页

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 

基　　金：National Natural Science Foundation of China [61403127, 61703144] Key Project of Henan Science and Technology, China 

主　　题：High impedance fault variational mode decomposition intrinsic mode function Teager-Kaiser energy operators time entropy 

摘      要：The focus of the paper is the difficulty of high impedance fault (HIF) detection in distribution network, and its ease to be confused with capacitor switching (CS) and load switching (LS). Based on the intermittent reignition and extinction characteristics of HIF current, this paper proposes a novel HIF detection method, which combines variational mode decomposition (VMD) and Teager-Kaiser energy operators (TKEOs). The HIF detection method is as follows: First, perform the VMD on transient zero sequence currents to obtain the intrinsic mode functions (IMFs) and select the IMFs with the largest kurtosis value as the characteristic IMFs. Second, calculate the characteristic IMFs to obtain TKEOs and divide into subintervals of TKEOs waveform to calculate the time entropy values. Finally, construct HIF detection criterion as follows: when time entropy value is 0, it is judged as CS or LS. When the entropy value is not 0, it is judged as HIF. A large number of simulations and field data tests show that the method is accurate and stable, and under the interference of 1 dB strong noise, it can accurately judge. Compared with other methods, the method has higher feature extraction accuracy, less calculation time, and better judgment accuracy.