Metodologia de avaliação da condição de operação de transformadores de distribuição por meio de análise da degradação térmica, integridade operacional e confiabilidade em ambiente de IoT

• Ano de defesa: 2021
• Autor(a) principal: Duarte, Leandro José
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Engenharia Elétrica
• Palavras-chave em Português: Análise de dados; Fator de envelhecimento equivalente; Índice de integridade; Redes elétricas inteligentes; Transformadores de distribuição; Data Analytics; Equivalent aging factor; Health index; Smart Grids; Distribution transformers; RPCA; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA; Transformadores elétricos
• Link de acesso: https://repositorio.ufu.br/handle/123456789/31841; http://doi.org/10.14393/ufu.di.2021.239

The emerging availability of data from Smart Grids, especially in the context of this work, related to the remote monitoring of distribution transformers, lacks methodologies capable of transforming this data into useful information for energy utilities. In this sense, this research proposes the use of three methodologies for assessing the condition of distribution transformers. The first explores the high correlation between currents, voltages, and temperatures, through Recursive Principal Component Analysis (RPCA), so that these quantities are synthesized as a point of operation in a two-dimensional map segmented in stress zones. A result is a visual tool for understanding the individual operating dynamics of transformers. The second proposed methodology uses several parameters to quantify the results obtained from electrical measurements, operational observations, and loading history, in an objective and quantitative index, capable of providing to some extent an indication of the general operating condition of the transformer. This index allows thousands of transformers from an electricity distribution company to be tabulated and ranked according to this proposed integrity index. In the third methodology, an analysis of the additional aging to which the transformer is being subjected is carried out, based on the operating conditions of this equipment, such as loading, tank temperature and environment, and physical characteristics of the transforming unit. Based on the results produced, this work contributes with methodologies that represent the basis for the evaluation of distribution transformers within an IoT system context capable of not only working with the monitoring of the electrical parameters of these assets but also convert this data into useful information that adds value to the IoT experience.