Localizing Data Manipulators in Distributed Mode Shape Identification of Power Systems

Abstract

In this paper we present two distributed algorithms for estimating the electro-mechanical oscillation mode shapes (eigen-vectors) of a power system using Synchrophasor measurements while also being aware of any cyber-threats that may bias these algorithms. We consider the power system to be divided into non-overlapping areas, each equipped with a local estimator. The local estimators exchange information for computing the mode shapes over a strongly connected communication graph, realized through an un-secure wide-area communication network (WAN). An attacker can intrude into this WAN, and manipulate the information exchanged between the estimators, thereby easily destabilizing the estimation loop. We develop mechanisms by which every estimator can either check the rank or inspect the singular values of appropriate data matrices. Any visible jump in the rank or singular values will enable the estimator to detect a potential manipulation. We validate our algorithms using a 4-machine 4-area power system and the IEEE 16-machine 68-bus system.