Particionamento de Sistemas de Medição para a Análise de Criticalidades na Estimação de Estado em Sistemas de Potência

Abstract

State Estimation (SE) is a tool that contributes to the safe and reliable operation of electrical power system networks. Its importance lies in the ability to determine the most likely operational state of the system, process the available measurements to mitigate the uncertainties inherent to the measurement process, and, notably, eliminate possible gross errors. This paper addresses the problem of analyzing measurements' criticalities, which are essential for indicating paths to the unobservability of the electrical network under supervision, brought about by possible measurement unavailability. This is a problem of a combinatorial nature, computationally tricky, with a solution within certain limits that depend on the size of the electrical network, the time required to solve the problem, the characteristics of the measurement system, and the value established for the maximum target cardinality of the analysis. Therefore, the focus is on the possibility of expanding such limits by adopting measurement system partitions that result in more palatable (tractable) instances for criticality analysis. Furthermore, this analysis is carried out through parallel processing. Numerical simulation results illustrate the application of the proposed partitioning methodology.