基于线性插值的异常采样值实时辨识方法

Real time identification of abnormal sampling data base on linear interpolation

采样数据异常会引起电力系统中继电保护装置和测控装置的误动作,现有方法在短数据窗下难以区分电气量异常采样值和故障时的正常采样值.本文推导了电力系统电气量相邻采样数据的线性插值余项误差,采样正常时该误差较小,出现异常采样点时该误差较大.在此基础上给出了误差的门槛值确定方法,以此来实时判断采样值是否异常.仿真验证了该方法的可行性.

The abnormal sampling data taken by electronic transformer may result in mal-operation in power system. It is different to discriminate between abnormal sampling data and normal data in fault period under short data window. The error between the adjacent remainder of Lagrange Linear Interpolation of power system is deduced secondly; the analysis point out that the error is small when the sampling data is normal and the error is big when the sampling data is abnormal; the solving method for the error threshold are given last. Simulation proves the feasibility of this method.