Ferramenta automática para diagnóstico das operações de proteção de distância em linhas de transmissão

• Ano de defesa: 2021
• Autor(a) principal: Matheus Henrique Couto Bueno
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA ELÉTRICA; Programa de Pós-Graduação em Engenharia Elétrica; UFMG
• Palavras-chave em Português: Redes de Petri temporais coloridas; Redes de Petri; Correlação; Diagnóstico de falta; Proteção de sistemas elétricos; Proteção de distância; Engenharia elétrica; Energia elétrica - Proteção; Redes de petri; Sistemas elétricos de potência
• Link de acesso: http://hdl.handle.net/1843/38663

New technologies related to automation and computer intelligence have caused the increase in complexity in power systems. Although those technologies help and facilitate the operation of power networks, enabling a greater situational awareness for system operators, they can introduce a plethora of new challenges to power utilities. During a fault, the engineering team can be flooded and overloaded with data from circuit telemetry, posing a difficulty for the power system state comprehension, the operational decision making and evaluation of the event. In this thesis a methodology is proposed for assessing distance protection relay operation in transmission lines, enabling the simplification of post-fault analysis and event reporting to the regulatory agencies. This approach is developed using a model-based technique with consistency reasoning to diagnose the system, built upon the simulation of a reference behavior for the equipment and the comparison of its results with data from the field, in search for inconsistencies. To enable this approach, the distance protection dynamic behavior has been reproduced using Colored Time Petri Nets (CTPN), a graphic and mathematical tool commonly used for concurrent, asynchronous, distributed and/or stochastic processes. A waveform comparison based on an assessment score calculation, combining different techniques for signal similarity determination, such as cross-correlation and temporal displacement among signals, is also proposed. The proposed methodology is able to evaluate multiple waveforms of on/off signals in intelligent electronic devices (IEDs) while granting robustness from system parameters or configurations variations. It was validated through electromagnetic transient simulations using the Alternative Transient Program (ATP) and IEEE’s proposed network for relay model testing. To prove its capacities in evaluating real occurrences, it was used to analyze two events occurred in a transmission utility, presenting correct results.