<?xml version="1.0" encoding="UTF-8"?>
<feed xmlns="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <title>TEDE Coleção:</title>
  <link rel="alternate" href="https://tedebc.ufma.br/jspui/handle/tede/525" />
  <subtitle />
  <id>https://tedebc.ufma.br/jspui/handle/tede/525</id>
  <updated>2026-04-13T15:31:41Z</updated>
  <dc:date>2026-04-13T15:31:41Z</dc:date>
  <entry>
    <title>Segmentação de acumulações de gás natural em imagens de reflexão sísmica através de técnicas 1D, 1.5D e 2D</title>
    <link rel="alternate" href="https://tedebc.ufma.br/jspui/handle/tede/6760" />
    <author>
      <name>MELLO, Henrique Ribeiro de</name>
    </author>
    <id>https://tedebc.ufma.br/jspui/handle/tede/6760</id>
    <updated>2026-02-05T13:25:53Z</updated>
    <published>2025-10-10T00:00:00Z</published>
    <summary type="text">Título: Segmentação de acumulações de gás natural em imagens de reflexão sísmica através de técnicas 1D, 1.5D e 2D
Autor: MELLO, Henrique Ribeiro de
Primeiro orientador: PAIVA, Anselmo Cardoso de
Abstract: The most common indirect identification method used by the oil and natural gas industry&#xD;
in the search for natural gas accumulations is conducting a survey based on the principle&#xD;
of seismic reflection. Specialists look for sudden changes in signal intensity, which may&#xD;
indicate the presence of natural gas, oil, or other hydrocarbons. In this thesis, we propose&#xD;
three methods to automate this task using seismic reflection data: one method that uses&#xD;
an Ensemble of U-Nets, a well-established architecture in image segmentation, and two&#xD;
others that make use of the Robustly Optimized Bidirectional Encoder Representation&#xD;
from Transformers Pretraining Approach (RoBERTa), a method originally developed in&#xD;
the context of natural language modeling. The methods are developed in such a way&#xD;
that we simulate what happens in practice, where models are trained in a given region&#xD;
and are used to evaluate unkown areas. To do so, we propose a train-test split policy&#xD;
for the public dataset used that simulates the use of the methods in a new region. The&#xD;
U-Net based method requires the annotation to be considerably larger than the original&#xD;
and is able to determine the annotated region more precisely. The method based on the&#xD;
RoBERTa architecture shows more false positives in areas of high acoustic-impedance&#xD;
&#xD;
variation, but it is capable of identifying possible smaller accumulations, as well as de-&#xD;
lineating much of the structure that stores the natural gas. All three methods present&#xD;
&#xD;
segmentation evaluation metrics superior to the state-of-the-art when considering the&#xD;
model’s generalization capability using a single dataset. On the F3 block dataset in the&#xD;
Netherlands, the methods achieved F1-scores between 0.36 and 0.62 on the proposed test&#xD;
set, while sensitivity varies between 0.65 and 0.79.
Instituição: Universidade Federal do Maranhão
Tipo do documento: Tese</summary>
    <dc:date>2025-10-10T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>CONTROL OF INTERNAL LOOPS IN CONVERTERS OPERATING AS VIRTUAL SYNCHRONOUS MACHINES FOR IMPROVED DYNAMIC RESPONSE</title>
    <link rel="alternate" href="https://tedebc.ufma.br/jspui/handle/tede/6591" />
    <author>
      <name>SERRA, Amiron Wolff dos Santos</name>
    </author>
    <id>https://tedebc.ufma.br/jspui/handle/tede/6591</id>
    <updated>2025-11-05T18:10:38Z</updated>
    <published>2025-10-03T00:00:00Z</published>
    <summary type="text">Título: CONTROL OF INTERNAL LOOPS IN CONVERTERS OPERATING AS VIRTUAL SYNCHRONOUS MACHINES FOR IMPROVED DYNAMIC RESPONSE
Autor: SERRA, Amiron Wolff dos Santos
Primeiro orientador: RIBEIRO, Luiz Antonio de Souza
Abstract: A crescente integração de fontes renováveis de energia aos sistemas elétricos modernos,&#xD;
impulsionada por preocupações ambientais e pela busca global por soluções sustentáveis, tem&#xD;
transformado significativamente o panorama da geração de energia. Nesse cenário, os recursos&#xD;
baseados em inversores tornaram-se essenciais para a conexão dessas fontes à rede elétrica. Tais&#xD;
recursos geralmente são controlados por estratégias do tipo seguidores de rede ou formadores de&#xD;
rede. Enquanto os inversores seguidores de rede dependem de malha de detecção de fase e&#xD;
requerem uma fonte externa de tensão para sincronização, os inversores formadores de rede são&#xD;
capazes de estabelecer de forma autônoma referências de tensão e frequência, sendo, portanto,&#xD;
adequados para operação tanto conectada à rede quanto isolada.&#xD;
Dentre as estratégias de formadores de rede, destacam-se as máquinas síncronas virtuais por&#xD;
sua capacidade de emular o comportamento inercial e de amortecimento dos geradores síncronos&#xD;
convencionais. Essa característica permite uma resposta dinâmica mais robusta em sistemas com&#xD;
alta penetração de fontes renováveis e baixa inércia. Inversores baseados em máquinas síncronas&#xD;
virtuais são capazes de sustentar a regulação de tensão e frequência, tanto em operação autônoma&#xD;
quanto em coordenação com a rede principal.&#xD;
Esta tese propõe o projeto e o aprimoramento das malhas internas de controle de um inversor&#xD;
baseado em máquina síncrona virtual, com o objetivo de melhorar a estabilidade do sistema&#xD;
elétrico e a resposta dinâmica. Especificamente, é desenvolvido um controlador de corrente com&#xD;
alta largura de faixa, capaz de atenuar a ressonância do filtro LCL e melhorar a rejeição a&#xD;
perturbações. Também é proposto um controlador de tensão com desacoplamento da entrada de&#xD;
perturbação, visando aprimorar a resposta dinâmica da máquina síncrona virtual. Com as malhas&#xD;
de controle internas projetadas, a máquina síncrona virtual apresenta desempenho dinâmico&#xD;
aprimorado quando conectada à rede, mesmo em uma ampla faixa de razões de curto-circuito,&#xD;
assim como durante variações de carga em modo isolado. A máquina síncrona virtual também&#xD;
demonstra melhor desempenho quando operando em uma microrrede, tanto em coordenação com&#xD;
um inversor seguidor de rede quanto em paralelo com outra máquina síncrona virtual.&#xD;
De modo geral, esta tese busca aprimorar as malhas internas de uma máquina síncrona&#xD;
virtual, permitindo que ela atue como uma estratégia eficaz de controle de potência em sistemas&#xD;
elétricos cada vez mais dominados por recursos baseados em inversores.
Instituição: Universidade Federal do Maranhão
Tipo do documento: Tese</summary>
    <dc:date>2025-10-03T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>Otimização da manutenção centrada na confiabilidade em redes de distribuição de energia elétrica: uma abordagem orientada a regulação baseada em desempenho</title>
    <link rel="alternate" href="https://tedebc.ufma.br/jspui/handle/tede/6523" />
    <author>
      <name>COSTA, Bruno Adrian Silva</name>
    </author>
    <id>https://tedebc.ufma.br/jspui/handle/tede/6523</id>
    <updated>2025-10-02T12:55:15Z</updated>
    <published>2025-09-05T00:00:00Z</published>
    <summary type="text">Título: Otimização da manutenção centrada na confiabilidade em redes de distribuição de energia elétrica: uma abordagem orientada a regulação baseada em desempenho
Autor: COSTA, Bruno Adrian Silva
Primeiro orientador: RODRIGUES, Anselmo Barbosa
Abstract: This study explores the Reliability-Centered Maintenance (RCM), supported&#xD;
by Predictive Reliability Analysis (PRA), focusing on optimizing maintenance&#xD;
frequency in power distribution networks, considering the strict regulatory context&#xD;
imposed by the National Electric Energy Agency (ANEEL). ANEEL oversees the&#xD;
distribution utilities by assessing individual and collective indicators related to the&#xD;
&#xD;
continuity of electricity supply. Consequently, utilities face penalties for non-&#xD;
compliance with these indicator targets, and an effective approach to mitigate these&#xD;
&#xD;
penalties involves careful planning of network equipment maintenance activities to&#xD;
maximize reliability indices and minimize maintenance costs. In this context, this&#xD;
research aims to optimize maintenance frequency to minimize costs and penalties&#xD;
associated with the System Average Interruption Duration Index (SAIDI). The&#xD;
proposed RCM model is based on chronological PRA, employing the State&#xD;
Enumeration Analytical Method (SEAM) and state aggregation techniques in&#xD;
Markov Chains. Additionally, it integrates the use of Optimal Power Flow (OPF) in&#xD;
electric distribution networks with meshed topology. The component degradation&#xD;
process is modeled using the Weibull distribution. The nonhomogeneous Markovian&#xD;
model resulting from this distribution is converted into a homogeneous model&#xD;
&#xD;
through series stage modeling. A Particle Swarm Optimization is applied in a multi-&#xD;
criteria optimization model. The results indicate significant penalty reductions with&#xD;
&#xD;
a high cost-benefit ratio, highlighting the economic efficacy of the proposed method&#xD;
for electric utilities, in alignment with ANEEL regulations.
Instituição: Universidade Federal do Maranhão
Tipo do documento: Tese</summary>
    <dc:date>2025-09-05T00:00:00Z</dc:date>
  </entry>
  <entry>
    <title>Quantidade mínima e alocação-regulação ótima de nós sensores móveis para uma cobertura circular completa da RSSF híbrida</title>
    <link rel="alternate" href="https://tedebc.ufma.br/jspui/handle/tede/6505" />
    <author>
      <name>SOUSA, Monik Silva</name>
    </author>
    <id>https://tedebc.ufma.br/jspui/handle/tede/6505</id>
    <updated>2025-09-22T17:48:05Z</updated>
    <published>2025-09-10T00:00:00Z</published>
    <summary type="text">Título: Quantidade mínima e alocação-regulação ótima de nós sensores móveis para uma cobertura circular completa da RSSF híbrida
Autor: SOUSA, Monik Silva
Primeiro orientador: FONSECA NETO, João Viana da
Abstract: Coverage is a measure of quality of service in wireless sensor networks (WSNs), and is&#xD;
 essential for monitoring a delimited region. This work addresses the dynamic coverage&#xD;
 problem of a hybrid WSN, proposing a methodology consisting of two complementary&#xD;
 methods. The first method aims to determine the minimum quantity of mobile sensor&#xD;
 nodes required to ensure full coverage of the monitoring region. This approach is based on&#xD;
 a theoretical framework composed of assumptions, propositions, theorems, and corollaries,&#xD;
 considering the sensor detection circles, the monitored region, and the agent displacement&#xD;
 circle. The model ensures full network connectivity and presents flexibility in dealing with&#xD;
 agent failures. The second method proposes an optimal DLQT-DLQR coverage control&#xD;
 scheme, based on the discrete algebraic Riccati equation (DARE), with the objective of&#xD;
 optimizing the displacement and regulation of dynamic nodes in their target positions.&#xD;
 The methodology ensures coverage without holes (gaps) and without connectivity losses&#xD;
 among nodes, minimizing the displacement time of mobile agents and avoiding collisions,&#xD;
 due to the preservation of the order of nodes along the displacement circle. The monitored&#xD;
 environment is modeled as a circular region, with a static sensor node in the center,&#xD;
 while the dynamic sensor nodes are allocated in a circumference with a radius smaller&#xD;
 than the coverage region. The proposed solution allows the efficient monitoring of hostile&#xD;
 environments, especially aquatic ones, contributing to environmental preservation, reducing&#xD;
 costs with idle sensors, and increasing the operational efficiency of the WSN. A comparative&#xD;
 analysis with the method based on Voronoi diagrams is performed, highlighting the need&#xD;
 for algorithms capable of calculating the desired positions of the agents to ensure flawless&#xD;
 coverage. The proposed method meets this requirement by determining the target positions&#xD;
 of the agents. In addition, the influence of the variation of the coverage radius on the&#xD;
 number of mobile sensor nodes for full monitoring is analyzed and a comparison is made&#xD;
 with a rectangular monitoring area. Finally, the performance of the proposed allocation&#xD;
 and regulation control is evaluated in relation to classical controllers, such as proportional&#xD;
 (P) and proportional integral (PI), in the time domain with figures of merit
Instituição: Universidade Federal do Maranhão
Tipo do documento: Tese</summary>
    <dc:date>2025-09-10T00:00:00Z</dc:date>
  </entry>
</feed>

