基于电流突变量曲线拟合的故障时刻检测算法

摘要

The traditional fault instant detection method based on the phase current increment is simple,but may exist a large detection error due to the influence of transient components.So with the current information collected by existing protection devices,the phase current increment is fitted and then the fault instant is obtained by inverse derivation.The transient process of short circuit in power system is analyzed,the variation law of phase current increment is explored and its generalized expression is deduced.The function curve of discrete phase current increment values is fitted by Prony algorithm and its characteristic parameters are obtained.The fault instant is further selected from the zero-crossing points of the fitting curve by improved waveform similarity algorithm.Simulation test is carried out in PSCAD/EMTDC to investigate the accuracy and influence factors of the fault instant detection,and the simulative results verify the feasibility and accuracy of the proposed algorithm.%传统基于相电流突变量的故障时刻检测方法虽然简单,但因受暂态分量的影响而存在较大的检测误差.利用现有保护装置采集的电流信息,对相电流突变量进行曲线拟合,并经逆向推导求出故障时刻.对电力系统的短路暂态过程进行分析,探究相电流突变量的变化规律并推导其广义表达式.利用Prony算法对离散的相电流突变量值进行函数拟合,得到表征其函数曲线的特征参数;通过改进现有的波形相似算法,进一步从拟合函数的过零点中筛选出故障时刻.结合PSCAD仿真测试,探究故障时刻的检测精度及其影响因素,仿真结果验证了所提算法的可行性和准确性.