Mestrado em Engenharia Elétrica

URI Permanente para esta coleção

http://repositorio.ufes.br/handle/10/17619

Nível: Mestrado Acadêmico
Ano de início:
Conceito atual na CAPES:
Ato normativo:
Periodicidade de seleção:
Área(s) de concentração:
Url do curso:

Navegar

Agora exibindo 1 - 5 de 375

Identificação de falhas em motores de indução trifásicos usando rede neural transformer convolucional
(Universidade Federal do Espírito Santo, 2025-03-24) Moraes, Vinicius Andrade Nunes de; Silva, Jair Adriano Lima ; http://lattes.cnpq.br/3099010533644898; Rocha, Helder Roberto de Oliveira; https://orcid.org/0000-0001-6215-664X; http://lattes.cnpq.br/8801325729735529; https://orcid.org/0009-0005-3889-354X; http://lattes.cnpq.br/1557334823568448; Encarnação, Lucas Frizera ; https://orcid.org/0000-0002-6162-7697; http://lattes.cnpq.br/5578918284508758; Augusto, Andre Abel; https://orcid.org/0000-0001-7171-3372; http://lattes.cnpq.br/2166068779417190
This work presents a hybrid approach using Convolutional Neural Networks (CNN) and Transformers for fault diagnosis in three-phase induction motors, focusing on the detection and classification of the severity of broken bar faults based on current and voltage signals. Electrical Signature Analysis (ESA), widely used in motor monitoring, offers several advantages. However, ESA-based techniques traditionally rely on spectral transformations, which can result in high computational cost and reduced generalization capability. CNNs can extract discriminative features directly from raw data, eliminating the need for preprocessing steps. The proposed study integrates CNNs with the attention mechanism of Transformers, which captures spatiotemporal dependencies in the data. The Convolutional Transformer Neural Network (CTNN) achieved approximately 97% accuracy when using the entire dataset, significantly outperforming classical machine learning algorithms such as Random Forest and k-Nearest Neighbors (KNN), which obtained 90% and 86% accuracy, respectively. The CNN, tested under similar conditions, achieved 96% accuracy. Compared to other methodologies involving multiple preprocessing steps and transformations to the frequency domain, the proposed approach achieves similar results, close to 100% accuracy, while being simpler, more efficient, and with greater generalization capability. Additionally, the methodology employs a reduced sampling rate, approximately six times lower than the original sampling rate of the dataset, contributing to computational cost reduction without compromising performance
Inversores fotovoltaicos multifuncionais aplicados na correção do fator de potência e na compensação de harmônicos
(Universidade Federal do Espírito Santo, 2025-03-12) Paiva, Fabrício Nunes; Medina, Augusto César Rueda; https://orcid.org/0000-0002-4291-3153; http://lattes.cnpq.br/7397584412509839; https://orcid.org/0009-0000-4598-0450; http://lattes.cnpq.br/8594725056951662; Antunes, Hélio Marcos André; https://orcid.org/0000-0001-8247-6448; http://lattes.cnpq.br/7601860538588447; Encarnação, Lucas Frizera; https://orcid.org/0000-0002-6162-7697; http://lattes.cnpq.br/5578918284508758; Ferraz, Rafael Santos Freire; https://orcid.org/0000-0001-8857-011X; http://lattes.cnpq.br/5323068276181437
The panorama of electricity systems around the world is undergoing a significant transfor mation, driven by the growing integration of alternative and renewable energy sources, such as solar, wind, and small hydroelectric plants (SHPs), among others. This phenomenon is widely known as distributed generation (DG). In Brazil, the photovoltaic sector stands out as a leader in plant installations, fueled by factors such as the widespread adoption of this technology, ease of installation, and the abundant solar radiation available in much of the country. However, integrating photovoltaic generation into electrical systems can pose challenges related to electric power quality, particularly with respect to the power factor. The generation of electricity by photovoltaic systems can lead to a decrease in the installation’s power factor, as the active power supplied to the load by the grid is reduced, while the reactive power remains unchanged compared to a system without photovoltaic generation. Another issue that can arise is the generation of harmonic cur rents caused by non-linear loads, which can compromise the integrity of equipment and conductors, as well as negatively affect the overall performance of the electrical system. In this context, the present paper proposes a methodology for improving power quality in a three-phase electrical system, utilizing a multifunctional converter based on a three-phase photovoltaic inverter with an LC filter (passive filter with inductor and capacitor). This system integrates both linear and non-linear loads and distributed solar generation. The approach aims to mitigate the negative impacts of low power factor and harmonic currents, optimizing the operation and reliability of the system. With the proposed multifunctional inverter, the power factor (lagging) was improved from 0.80 to 0.92, while harmonic current distortion was reduced from 15% to 3.34%. Additionally, another advantage observed in this work was the reduction in the number of proportional-resonant (PR) controllers in the control loop using dq0-coordinate control, with a single PR controller capable of controlling two harmonic frequency orders in the abc coordinates. The control strategy was validated through computer simulations implemented in the development environment of the MATLAB/Simulink® software
Otimização metaheurística e aprendizado de máquina para identificação da doença de Parkinson por sinais de voz
(Universidade Federal do Espírito Santo, 2025-01-28) Garcez, Peter Gleiser; Salles, Evandro Ottoni Teatini; Ciarelli, Patrick Marques; https://orcid.org/0000-0003-3177-4028; http://lattes.cnpq.br/1267950518719423; Krohling, Renato Antônio; Komati, Karin Satie
Approximately 220,000 Brazilians have Parkinson’s Disease (PD), which affects 1% to 3%of the world’s population over 65 years, according to WHO estimates. PD causes a continuous and gradual loss of dopamine-producing neurons, a neurotransmitter essential for muscle function performance, especially speech motor control, causing impairment in voice quality. This study aims to implement feature selection and machine learning hyperparameter tuning through optimization metaheuristics to identify PD using features extracted from voice signals. At first, the metaheuristic Adaptive Hybrid-Mutated Differential Evolution (A-HMDE) is applied to select features from the Parkinson’s Disease Classification dataset, consisting of 752 features extracted from 756 voice signal samples. Next, considering the selected features, we tuned the hyperparameter of the Random Forest (RF) and k-Nearest Neighbors (kNN) models, as well as of the Convolutional Neural Network 1D (CNN 1D) model using metaheuristic. A reduction from 752 to 75 features was achieved, representing a selection rate of less than 10%, with an accuracy of 91.63% and a recall of 99.39% obtained by the RF classifier. The results demonstrate the effectiveness of the metaheuristics used for identifying Parkinson’s Disease through voice, and the need to develop datasets with unprocessed vocal signals to explore the performance of convolutional networks operating on raw signals for PD classification
Aplicação da técnica Motor Current Signature Analysis (MCSA) em motores de uma mesa de resfriamento de rolos
(Universidade Federal do Espírito Santo, 2024-09-10) Chagas, Rafael Mariano; Santos, Walbermark Marques dos ; https://orcid.org/0000-0002-9871-6028; http://lattes.cnpq.br/5558697161842579 ; https://orcid.org/0009-0000-4045-2181; http://lattes.cnpq.br/4155218335098440; Rocha, Helder Roberto de Oliveira ; Encarnação, Lucas Frizera ; Barcelos, Silvangela Lilian da Silva Lima ; Nascimento, Thais Pedruzzi do; Nascimento, Thais Pedruzzi do
Electric current signature analysis (ESA) has been widely used as a solution for projects to improve operational reliability in industries, since there are many studies that prove its effectiveness. The variability of ESA conditions is a set of techniques capable of detecting failures by reading electrical signals collected remotely. According to Bonaldi et al (2007), among the techniques that make up ESA, the MCSA (Motor Current Signature Analysis) technique is the most widely used technique in the industrial sector and stands out for its comprehensiveness and simplicity, since it only requires the analysis of the machine's current spectrum signal to detect abnormal conditions. In the search for better industrial development, the study of techniques that improve equipment availability is of utmost importance. Within this context, this dissertation aims to study the applicability of using this technique to improve the reliability of a roller table, with 264 electric motors installed in the process in a sequential manner, where any failure in one of these pieces of equipment impacts the entire process. These motors are driven by frequency inverters and this type of drive can pollute the MCSA response and therefore the problem will also be discussed. The results presented in this dissertation show that the fault detection technique works, even with all the practical realities encountered in the actual implementation of equipment
Comportamento da proteção de sobrecorrente em redes com geração distribuída operando em ilhamento intencional
(Universidade Federal do Espírito Santo, 2024-08-23) Paula, Toribio Cruvinel de; Simonetti, Domingos Sávio Lyrio ; https://orcid.org/0000-0001-5920-2932; http://lattes.cnpq.br/1107005171102255; https://orcid.org/0009-0002-2261-8517; http://lattes.cnpq.br/8073441410217732; Rueda Medina, Augusto César ; https://orcid.org/0000-0002-4291-3153; http://lattes.cnpq.br/7397584412509839; Có, Márcio Almeida ; https://orcid.org/0009-0001-6026-0125; http://lattes.cnpq.br/9674164201696461
Distributed generation (DG) has become a fundamental part of the modern electrical system, especially with the increasing adoption of renewable energy sources such as solar and wind. However, the integration of DG presents significant challenges for the protection of distribution systems, particularly in islanding situations, where the isolated operation of microgrids can compromise the coordination of protection devices. Despite advancements in protection technology, there are still gaps in understanding how DG affects the coordination of overcurrent protections, resulting in operational failures and safety risks. Existing automated solutions have not proven to be sufficiently effective in addressing the complexities introduced by DG, especially in islanded operation scenarios. This work proposes a detailed analysis of the impacts of islanding in microgrids with GDFV from the perspective of overcurrent protection. The methodology involves simulations in a representative environment that incorporates industrial loads and photovoltaic generation, using MATLAB/Simulink software to model the behavior of protection devices under different operating conditions. The results indicate that the integration of distributed photovoltaic generation requires a reassessment and adjustment of existing protection systems to ensure the continuity and safety of distribution network operations. The implementation of adaptive technologies and the strategic use of new protection functions are essential to address the challenges posed by this new energy configuration

Navegar

Submissões Recentes