Uma nova abordagem para correção de erros grosseiros em estimação de estados utilizando algoritmo genético

Jeanderson Soares Mingorança; Igor Delgado de Melo; Matheus Pereira Antunes; Braulio César de Oliveira; Matheus Alberto de Souza

doi:10.20906/sbse.v2i1.2921

Jeanderson Soares Mingorança Faculdade de Engenharia Elétrica, Universidade do Federal de Juiz de Fora, MG
Igor Delgado de Melo Faculdade de Engenharia Elétrica, Universidade do Federal de Juiz de Fora, MG
Matheus Pereira Antunes Faculdade de Engenharia Elétrica, Universidade do Federal de Juiz de Fora, MG
Braulio César de Oliveira Faculdade de Engenharia Elétrica, Universidade do Federal de Juiz de Fora, MG
Matheus Alberto de Souza Faculdade de Engenharia Elétrica, Universidade do Federal de Juiz de Fora, MG

DOI: https://doi.org/10.20906/sbse.v2i1.2921

Keywords: State Estimation, WLS, PMUs, Smart Meters

Abstract

This paper presents a new methodology for correcting gross errors in power distri- bution systems state estimation (SE). A SE algorithm is developed based on the Weighted Least Squares (WLS) using a hybrid approach in which non-synchronized measurements and PMUs (Phasor Measurement Units) are used together. At the end of the process, the Normalized Residuals (NRs) are calculated for each measurement and the existence of gross errors is detected based on statistical analysis of the NRs. Suspicious measurements are identified by the highest values of NRs. A calibration factor is associated with each suspected measurement to estimate the system state and correct the gross errors. In order to determine the exact values of the calibration factors, a new formulation of an optimization problem is proposed, being the solution provided by genetic algorithm. Simulations are conducted using the IEEE 33-bus test system to validate the proposed method proving its efficiency and robustness.