Localização automática de faltas em linhas de transmissão de circuitos duplos

• Ano de defesa: 2007
• Autor(a) principal: Thiago Matozinhos de Souza
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: corrente; tensão; Circuito; Engenharia elétrica
• Link de acesso: http://hdl.handle.net/1843/BUOS-8CYGS9

Fault location on parallel lines still has an inferior level of development ifcompared to the obtained progress on single circuit lines. The fault location forn double circuit or parallel lines can be considerably influenced by the mutual coupling between circuits. The majority of methods for fault location were not developed considering the existence of the second circuit. In such way, there is an incomplete modeling, which leads the algorithm to a wrong faults incidence point estimativewhen used on a double circuit line. In other cases, a very simple lines model is used, that also leads to a locator with undesirable precision, especially when dealing with relatively long lines, typically found in Brazil.This dissertation will present a study of an automatic fault location technique on double circuit transmission lines, with the use of voltage and current data from two terminals of both circuits. The developed methodology uses matrix decomposition that transforms double circuit lines into two decoupled circuits, called additive circuit and differential circuit. The next step consists on using single circuit Johns & Jamallis method on differential circuits data. It is an original approach inthe frequency domain. Beyond faults incidence point, the Unreliable Factor will also be presented; witch is another parameter that lets infer the quality of the results. The developed routines had been implemented through Matlab software. The fundamental characteristics of the fault location algorithm was determined by testing the data generated by ATP, a transient simulator program. A wide range of line parameters was evaluated such as fault position, type, and resistance; sourceimpedance ratio and others. A real data case was also used. The results showed that the developed methodology is able to produce a better fault location estimative when compared to the original Johns & Jamallissingle circuit method with the non-faulted circuit ignored.