Topological Approach for Identifying Critical Measurements and Sets in State Estimation

Abstract

Power system state estimation (SE) is an energy management system application responsible for providing a consistent real-time database, instrumental in monitoring the system. SE input data are redundant observations (measurements) of the system state (complex bus voltages) taken in a given network configuration. Measurement redundancy is an essential requirement for the SE results' reliability, determined by the quantity, location, and type of measurements received for processing. There are two ways to approach the observability/ criticality analysis, namely topological (a graph theory-based) and numerical (performed by arithmetic operations on matrices). From the conceptual viewpoint, topological methods are adequate since the problem in question is considered structural-natured, dependent on the network topology and the type/placement of measurements. Criticality analysis has been considered vital to reveal the different network observability degrees established by the measuring system. The most severe conditions are related to the occurrence of single critical measurements and critical sets of measurements. These conditions refer to imminent network unobservability and SE limitation to detect/identify the presence of spurious measurements. This paper proposes a graph theory-based method devoted to identifying essential elements to SE. Conventional measurements (branch power flows and bus power injections) coming from SCADA systems, as well as synchrophasors (phase angles and branch current measurements), are considered. Simulation results carried out on the IEEE 14-bus test system are provided.