利用行波固有频率和原子能量熵的故障选相方法

A Fault-Phase Selection Method Based on Natural Frequency of Traveling Wave and Atomic Energy Entropy

准确识别故障信号中所含固有频率成分特征是基于该量的选相算法的核心,分析故障时相电流中所含固有频率的能量与故障类型的关系,提出一种直接分析相电流中固有频率特征的选相方法;多相短路故障时,在滤除工频分量后,各故障相中均含有显著的固有频率且频率相等,非故障相中的固有频率能量很弱:单相短路时各相均难有稳定的固有频率。应用原子分解法分析故障电流,若固有频率成分显著存在,则其对应主导原子且能量熵值最大,根据相电流中主导原子的能量熵值及其频率即可进行选相。经仿真验证,文中方法具有较理想的固有频率辨识能力,选相结果稳定,可靠。

Accurately recognizing the characteristics of natural frequency component contained in fault signal is the core of the fault-phase selection algorithm based on the component.The relation between the fault type and the natural frequency energy contained in phase current during the fault is analyzed,and a fault-phase selection method by directly analyzing the natural frequency characteristics in phase current is proposed based on the following phenomenon：during multi-phase short-circuit fault,when power frequency component in fault current is filtered,in all faulty phases there are obvious natural frequency components and these natural frequencies are the same while the natural frequency energy in the healthy phase is very weak; under single-phase grounding fault there is not stable natural frequency components in all phases.The atom-decomposition algorithm is applied in fault current analysis,if there is obvious natural frequency component,this component corresponds to the dominant atom and its energy entropy is the maximum,thus according to the energy entropy value of dominant atom in phase current and the natural frequency,the fault-phase can be selected.Simulation results show that the proposed method possesses satisfied ability to recognize natural frequency and the results of fault-phase selection are stable and reliable.