Vulnerability of Largest Normalized Residual Test and <(b)over cap> - Test to Gross Errors

Autor(es)

Massignan, Julio A. D. ; London Jr, Joao B. A. ; Vieira, Camila S. ; Miranda, Vladimiro ; IEEE

• Tipo de documento: Artigo Científico
• Fonte: 2020 IEEE POWER & ENERGY SOCIETY GENERAL MEETING (PESGM); v. N/A, p. 5-pg., 2020-01-01.

Resumo

Power systems rely on a broad set of information and sensors to maintain reliable and secure operation. Proper processing of such information, to guarantee the integrity of power system data, is a requirement in any modern control centre, typically performed by state estimation associated with bad data processing algorithms. This paper shows that contrarily to a commonly assumed claim regarding bad data processing, in some cases of single gross error (GE) the noncritical measurement contaminated with GE does not present the largest normalized residual. Based on the analysis of the elements of the residual sensitivity matrix, the paper formally demonstrates that such claim does not always hold. Besides this demonstration, possible vulnerabilities for traditional bad data processing are mapped through the Undetectability Index (UI). Computational simulations carried out on IEEE 14 and IEEE 118 test systems provide insight into the paper proposition. (AU)

• Processo FAPESP: 16/19646-6 - Estimador de estado trifásico multiárea para sistemas de distribuição de larga escala
• Beneficiário: Julio Augusto Druzina Massignan
• Modalidade de apoio: Bolsas no Brasil - Doutorado