Uma proposta para a atribuição de responsabilidades sobre desequilíbrios de tensão através da análise de componentes independentes

• Ano de defesa: 2022
• Autor(a) principal: Moraes, Márcio Arvelos
• Tipo de documento: Tese
• 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 Componentes Independentes Complexas; Complex Independent Component Analysis; Compartilhamento de Responsabilidades sobre Desequilíbrios de Tensão; Sharing Responsibility for Voltage Imbalances; Qualidade da Energia Elétrica; Power Quality; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA::SISTEMAS ELETRICOS DE POTENCIA; Engenharia elétrica; Tensão - Análise de componentes principais; Energia elétrica - Distribuição; Sistemas de energia elétrica - Estabilidade
• Link de acesso: https://repositorio.ufu.br/handle/123456789/36600; http://doi.org/10.14393/ufu.te.2022.5040

Due to the occurrence of different phenomena associated with deviations in the power quality of electrical supplies, among the challenges inherent to the subject, efforts have been undertaken to establish means for determining the contributions and origin of the disturbances. Within this context, emphasizing the levels of imbalance of voltages present in the networks, the final results obtained for an electrical installation are the composition effects linked to the agents` operation. This area of study is known as Responsibility Sharing and is presented as the core of this doctoral research. Within this theme, despite the recognition of several published scientific works, there is still a long way to go until the definition of procedures that prove to be consistent, feasible to implement in the field, and whose results offer the required reliability for the attribution of responsibilities between the parties. Thus, evaluative and comparative works are carried out on the existing main methods used to share responsibilities regarding imbalances in electrical systems, identifying their properties and limitations, and finally, the proposition of a non-invasive methodological strategy for the purposes set forth here. The established process is based on the Complex Independent Component Analysis (CICA) technique, using Higher Order Statistics (HOS), statistical independence, non-Gaussianity, and pre-processing methods as whitening. The strategy is presented in terms of its physical and mathematical foundations and its application direction for this research. Evaluative studies on the proposal's performance aim at the computational evaluation of the methodology regarding its practical feasibility and effectiveness in relation to the desired objectives. The results achieved show great prospects for using the CICA method to determine the contribution portions between the agents, thus contributing so that any mitigating measures can be established fairly between the agents.