Preditores aplicados na inicialização inteligente do método da soma de potências em série temporal

• Ano de defesa: 2019
• Autor(a) principal: Henriques, Jéssica Madruga de Miranda
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Engenharia Elétrica; Programa de Pós-Graduação em Engenharia Elétrica; UFPB
• Palavras-chave em Português: Fluxo de carga na série temporal; Análise de fluxo de carga; Preditores; Preditores inteligentes; Inicialização inteligente; Time series load flow; Load-flow analysis; Predictors; Intelligent predictors; And intelligent initialization; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/19688

The conventional analysis of the steady-state power systems is done executing the load-flow and we can analyze the behavior of the system. However, with the insertion of distributed energy, for example, the nature of the system becomes more intermittent. As a result of problems such as this and also because of a greater availability of processing and memory that allows us to perform actions that were previously unfeasible, the idea of executing a Load-Flow in the Time Series, which opens the possibility to analyze how the state of the system varies daytime, becomes desired. For this kind of analysis, many iterations in the load flow method are done for each time sample. In this masters final project, an algorithm was developed with the purpose of reducing the total number of iterations required in the load flow. For this, the analysis and implementation of predictors that make an intelligent initialization of the tensions in the network bars were implemented before executing the load flow. The number of iterations required for convergence was then compared with a reference value, obtained executing the load-flow without intelligent initialization. In addition to the three intelligent predictors found in the literature and with good optimization results, three new predictors were proposed. For this analysis, we used data from a real system of 63 bars with real load and four different scenarios of load variation to validate the predictors. The results presented proved the efficacy of the predictors with the intelligent initialization and also an improvement with the new proposed predictors.