Modelagem orientada a objetos de estimadores de estados aplicados à supervisão da segurança operativa de redes elétricas

Abstract

The use of advanced hardware and digital computers in electrical management systems potentially enhance the supervision and control tasks in real time operation mode. These functions allow the safe operation of electrical networks, which effectively improve the operator decision. Specifically, the state estimators applied to power systems have an important role along with the SCADA systems, in order to evaluate the network states as close as possible to the actual states of the monitored system. This work evaluates a robust state estimator algorithm performance when it is written in object-oriented programming language. The program makes use of the Iterative Weighted Least-squares method (IWLS) with equality constraints (EC). The employed methodology has the feature of processing bad data during the iterative search for the problem solution, even when it is extended to the state estimation problem with equality constraints. In this work, the equality constraints can be modeled through two approaches: a) Classical Weighting Method (WM); b) Weighting Method with Iterative Refinement (WMIR). Independently from the fact that regular and/or pseudo measurements have been classified as leverage points, the employed methodology, named IWLS-EC, allows the satisfactory processing of bad data and/or topology errors. The algorithm performances are explicitly tracked throughout the iterations carried out on a small example system of 5 buses. Moreover, the proposed methodology is also evaluated in an actual system in the south of Brazil with 126 buses, and also in other realistic systems with 340, 730 and 1916 buses. The obtained results are compared, analyzed and commented on in the chapter that describes the set of simulations carried out on test systems.