Exploring Machine Learning-Based Solutions for Fault Classification and Region Identification in Onshore Wind Farm Collector Systems

Autor(es)

Davi, Moises J. B. B. ; Cunha, Talita M. O. A. ; Oliveira, Emanuel P. G. ; Oleskovicz, Mario

• Tipo de documento: Artigo Científico
• Fonte: 2024 21ST INTERNATIONAL CONFERENCE ON HARMONICS AND QUALITY OF POWER, ICHQP 2024; v. N/A, p. 6-pg., 2024-01-01.

Resumo

With the increasing expansion of renewable energy resources, wind farms are becoming predominant, especially with the progress of Inverter-Based Resources (IBRs). Although IBRs are operationally effective, it is recognized that this generation type has directly impacted the operation of protection and fault diagnosis functions. In this context, this paper investigates the potential of Machine Learning (ML)-based solutions for classifying faults and identifying faulted regions in onshore wind farm collector systems. For the studies, a realistic wind farm system topology is modeled in PSCAD software, and fault scenarios are represented in the collector systems with different fault parameters and wind farm generation levels. These studies highlight the best-performing ML methods for fault diagnosis within onshore wind farms and provide insights on the best measurement points for successful fault classification and fault region identification. (AU)

• Processo FAPESP: 20/15230-5 - Centro de Pesquisa e Inovação de Gases de Efeito Estufa - RCG2I
• Beneficiário: Julio Romano Meneghini
• Modalidade de apoio: Auxílio à Pesquisa - Programa Centros de Pesquisa em Engenharia
• Processo FAPESP: 22/00483-0 - Metodologia para diagnóstico de faltas em linhas de transmissão com alta penetração de geração eólica interfaceada por inversores
• Beneficiário: Moisés Junior Batista Borges Davi
• Modalidade de apoio: Bolsas no Brasil - Doutorado