Avaliação da estabilidade de tensão utilizando análise modal e estimação de estados: uma abordagem considerando o impacto de erros de medição para a identificação de barras críticas

Abstract

This dissertation presents a new approach to the evaluation of voltage stability in power systems using modal analysis applied to the results obtained using the State Estimation (SE) technique. In the proposed methodology, measurements are obtained from continuously monitored electrical networks and these are processed using the Weighted Least Squares (WLS) method, which is modified from a hybrid approach in which synchronized and unsynchronized measurements are merged in order to provide greater robustness to the State Estimators. Thus, synchronized phasor meters are used together with smart meters and traditional SCADA (Supervisory Control and Data Acquisition) system sensors in order to determine the most likely operating state of a monitored power system. From the introduction of the concept of statistical robustness in the proposed method, the aim is to suppress the impact of gross errors, ensuring greater reliability to the results, using Pearson’s Chi-square and Normalized Maximum Residue (MRN) tests. From the estimated values of voltage magnitude and angles in all buses in the system, it is possible to calculate a sensitivity matrix of the reactive powers in relation to the estimated voltage magnitudes using algebraic equations. Modal analysis is applied to this matrix, determining its eigenvalues and eigenvectors which are used to estimate participation factors for each bus in the system, indicating the most critical in relation to the voltage stability problem. The 30 transmission bus system, the 33 medium voltage bus system and a university distribution system that models the electrical network of the UFJF (Federal University of Juiz de Fora) were used to conduct case studies in order to validate the method through statistical metrics. Furthermore, studies, analyzes and results about gross errors are presented, ways to detect them, identify them and suppress/correct them to maintain the confidence of the SE and, consequently, ensure that voltage stability studies can be carried out in practice, reducing the negative impacts of possible measurement errors.