A hybrid method (MVMO+TSM) for the parameter estimation of one-axis generators model

• Ano de defesa: 2023
• Autor(a) principal: Vigoria, Edsson Frank Medina
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: convergence; convergência; estimação de parâmetros; gerador síncrono; hybrid method; método híbrido; MVMO; parameter estimation; Sensibilidade de Trajetória; synchronous generator; Trajectory Sensitivity
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/18/18154/tde-14062023-095148/

The knowledge of parameters of transient model of the synchronous generator is fundamental for dynamic studies, which is used to make planning and analysis of security of the electrical energy system. Normally, the parameters are determined through offline methods (with the generator disconnected from the grid), however, new research uses online methods (with the generator connected to the grid) due to the advantages to obtain parameter values in different operation conditions, with this purpose the method needs to be reliable and robust. In this work, a hybrid method is presented to estimate the parameters of the one-axis model of the synchronous generator with the purpose of making the estimation with measures that are easy to obtain and robust in relation to the convergence of parameters. This hybrid method is composed of a metaheuristic method called Media-Variance Mapping Optimization (MVMO) and a non-linear programming method called Trajectory Sensitivity (TSM). The MVMO method is used to provide a smart initial guess for the second method (TSM), which will be used to find the local minimum, therefore, the hybrid method takes advantage of the metaheuristic method and non-linear programming. The method is validated by estimating the parameters of a 2 kva synchronous generator in a small power system assembled in the laboratory with satisfactory results.