Advanced search
Start date
Betweenand

Analysis of artificial neural networks methodologies for event classification using Synchrophasors

Grant number: 17/25425-5
Support type:Scholarships in Brazil - Master
Effective date (Start): April 01, 2018
Effective date (End): May 31, 2020
Field of knowledge:Engineering - Electrical Engineering - Power Systems
Cooperation agreement: Coordination of Improvement of Higher Education Personnel (CAPES)
Principal researcher:Daniel Dotta
Grantee:Orlem Lima dos Santos
Home Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
Associated research grant:16/08645-9 - Interdisciplinary research activities in electric smart grids, AP.TEM
Associated scholarship(s):19/08200-5 - Identifying power system events using a long short-term memory neural network, BE.EP.MS

Abstract

In recent years, the electric power industry has been undergoing with strong modifications in its infrastructure with the pressure for solutions less harmful for the environment. The massive incorporation of renewable energy sources, such as wind and photovoltaic energy, characterized by intermittency and low inertia, increases the operational complexity and uncertainty of modern Electrical Energy Systems (EES). Additionally, in the Brazilian case, the significant increase of the HVDC (High Voltage Direct Current) transmission systems for the transfer of large blocks of energy to the southeastern region significantly increases the risk of critical contingencies. In this context, methods capable of improving the diagnosis capacity of operation in an EES, verifying its performance, are high relevance. The correct diagnosis and analysis of occurrences can contribute toward preventing possible problems such as shortages and even blackouts. Therefore, the use of new tools for monitoring, like the Wide Area Measurement Systems (WAMS), is of great importance, because they enable the capture of the dynamic behavior of the electrical power systems and present themselves as a promising technology for the monitoring, control and security management of large systems. The WAMS networks are more and more increasingly used in industry and are considered as fundamental for the implementation of the concept of Smart Transmission Grids. The current challenge is to handle the large amount of data captured by the WAMS and process them in a way to extract relevant information which may give subsidies to the operator of the EES. In this regard, the exploitation of methods based on machine learning is a relevant topic of research. It appears that it already exists in the literature, several methods both statistical and deterministic that are used in the classification of events. The most relevant and seen in the literature among the statistical methods are the Artificial Neural Networks (ANN), which has been applied with great efficiency in various applications in the EES, including the classification of events. However, a large part of the research carried out so far has focused on the application of static neural networks that do not have the capacity to identify the non-stationary (or dynamic) characteristics present in an EES. So, with the purpose to extend the knowledge about these technologies for classification of events in EES, this Master's degree project aims to investigate and analyze the application of ANN in the classification of events using synchrophasors. For this, it proposes to implement a classifier based on dynamic neural networks in Matlab/Simulink and apply it in the classification of events already occurred in Brazilian Interconnected Power System (BIPS), obtained through the MedFasee project. (AU)

News published in Agência FAPESP Newsletter about the scholarship:
Articles published in other media outlets (0 total):
More itemsLess items
VEICULO: TITULO (DATA)
VEICULO: TITULO (DATA)

Academic Publications
(References retrieved automatically from State of São Paulo Research Institutions)
SANTOS, Orlem Lima dos. Aplicação de interpretabilidade para melhorar o desempenho de um classificador LSTM para eventos de sistema de potência. 2020. Master's Dissertation - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação.

Please report errors in scientific publications list by writing to: cdi@fapesp.br.