The usage of median filtering in power system state estimation

摘要

State estimation (SE) is a widely used tool in power system control centers. It improves accuracy of remote measurements enabling direct power system state monitoring. In the article, the usage of median filtering (MF) in SE algorithms is proposed. Modeling of SE by MF and weighted least square method (WLS) was performed under conditions of the most challenging situations for SE, such as communication channel breakdown and current transformer saturation. The results of the modeling revealed that MF does not concede in accuracy to WLS method and in some cases even turns out to be a more effective mean. It should be noted that the simplicity of MF allows reducing of computational efforts in comparison with WLS. Thus, MF can be used in dynamic power systems as approximate estimation.