Rafael Lima de Carvalho

Rafael Lima de Carvalho

Auto-generated profile, based on publicly available information

Rafael Lima de Carvalho

Universidade Federal do Tocantins

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
31 December 2021 in IEEE Latin America Transactions
Verified
Journal Article
Closed access
ASAdelicio Maximiano Sobrinho
Adelicio Maximiano Sobrinho
JCJosé Roberto Camacho
José Roberto Camacho
RCRafael Lima de Carvalho
Rafael Lima de Carvalho
SSSergio Manuel Rivera Sanhueza
Sergio Manuel Rivera Sanhueza
SFStefani Freitas
Stefani Freitas
The aim of this research is to present the studies carried out to design three-phase core type distribution transformers, with the aid of optimization techniques. The minimization of losses is presented in two steps, first through the use of a mono-objective function, and second with the use of multi-objective function to minimize losses and the total mass of the active part of the transformer. The algorithms used are: Differential Evolution and Particle Swarm Optimization, their performances are compared through the results obtained. The main project parameters, such as core dimensions, total losses, no load current and the energizing current are estimated analytically through OCTAVE software. The analysis of magnetic flux density in the core is simulated using the Finite Element Method. Multiobjective optimization allows working with two or more conflicting objectives, and at each iteration, it stores the various non-dominant Pareto Front solutions, helping designers to choose the solution that best meets their needs. The results obtained with the mono-objective and multiobjective optimization techniques were interesting to minimize the losses and/or cost of the project

19 February 2022 in One Health
Verified
Journal Article
Open access
DPDavid Nadler Prata
David Nadler Prata
WRWaldecy Rodrigues
Waldecy Rodrigues
DTDaniela Trevisan
Daniela Trevisan
WCWainesten Camargo
Wainesten Camargo
HFHumberto Frizzera
Humberto Frizzera
PBPaulo Henrique De Souza Bermejo
Paulo Henrique De Souza Bermejo
This study investigates the spatial differences in the occurrence of COVID-19 in Brazilian Tropical Zone and its relationship with climatic, demographic, and economic factors based on data from February 2020 to May 2021. A Linear Regression Model with the GDP per capita, demographic density and climatic factors from 5.534 Brazilian cities with (sub)tropical climate was designed and used to explain the spread of COVID-19 in Brazil. The model shows evidence that economic, demographic and climate factors maintain a relationship with the variation in the number of cases of COVID-19. The Köppen climate classification defines climatic regions by rainfall and temperature. Some studies have shown an association between temperature and humidity and the survival of SARS-CoV-2. In this cohort study, Brazilian cities located in tropical regions without a dry season (monthly rainfall > 60 mm) showed a greater prevalence than in cities located in tropical regions with a dry season (some monthly rainfall < 60 mm). Empirical evidence shows that the Brazil's tropical-climate cities differ in the number (contamination rate) of COVID-19 cases, mainly because of humidity. This study aims to alert the research community and public policy-makers to the trade-off between temperature and humidity for the stability of SARS-COV-2, and the implications for the spread of the virus in tropical climate zones.

10 June 2020 in Academic Journal on Computing, Engineering and Applied Mathematics
Verified
Journal Article
Open access
PCPedro Henrique Cardoso Camelo
Pedro Henrique Cardoso Camelo
RCRafael Lima De Carvalho
Rafael Lima De Carvalho
The Multilayer Perceptron (MLP) is a classic and widely used neural network model in machine learning applications. As the majority of classifiers, MLPs need well-defined parameters to produce optimized results. Generally, machine learning engineers use grid search to optimize the hyper-parameters of the models, which requires to re-train the models. In this work, we show a computational experiment using metaheuristics Simulated Annealing and Fast Simulated Annealing for optimization of MLPs in order to optimize the hyper-parameters. In the reported experiment, the model is used to optimize two parameters: the configuration of the neural network layers and its neuron weights. The experiment compares the best MLPs produced by the SA and FastSA using the accuracy and classifier complexity as comparison measures. The MLPs are optimized in order to produce a classifier for the MNIST database. The experiment showed that FastSA has produced a better MLP, using less computational time and less fitness evaluations.

1 August 2019 in Revista Observatório
Verified
Journal Article
Open access
PDF file
MRMauricio Santana Ribeiro
Mauricio Santana Ribeiro
RCRafael Lima de Carvalho
Rafael Lima de Carvalho
AOAry Henrique Oliveira
Ary Henrique Oliveira
GBGlenda Michele Botelho
Glenda Michele Botelho
WPWandson Mendes Pessoa
Wandson Mendes Pessoa
O crescimento populacional mundial e a respectiva concentração de grande parte destas pessoas nas áreas urbanas gera desafios para os gestores. Neste contexto, surgem as cidades inteligentes, que são definidas como um novo método para o desenvolvimento sustentável das cidades. Este artigo analisa as concepções conceituais quanto às cidades inteligentes comparado ao crescimento populacional urbano das cidades e discutindo se o processo de expansão territorial e a verticalização das cidades. A problemática desenvolvida procura discutir se de fato as cidades inteligentes são o caminho para o preparo do crescimento populacional. Os resultados mostram que há significativa hipótese para essa condição.

3 July 2026
Verified
Book Chapter
Closed access
PDF file
DSda Silva
da Silva
MMMarcos Dyeimison Moreira
Marcos Dyeimison Moreira
DJde Jesus
de Jesus
CSCarlos Silva
Carlos Silva
DSde Souza
de Souza
RLRafael Lima
Rafael Lima

10 June 2020 in Academic Journal on Computing, Engineering and Applied Mathematics
Verified
Journal Article
Open access
PCPedro Henrique Cardoso Camelo
Pedro Henrique Cardoso Camelo
RCRafael Lima De Carvalho
Rafael Lima De Carvalho
The Multilayer Perceptron (MLP) is a classic and widely used neural network model in machine learning applications. As the majority of classifiers, MLPs need well-defined parameters to produce optimized results. Generally, machine learning engineers use grid search to optimize the hyper-parameters of the models, which requires to re-train the models. In this work, we show a computational experiment using metaheuristics Simulated Annealing and Fast Simulated Annealing for optimization of MLPs in order to optimize the hyper-parameters. In the reported experiment, the model is used to optimize two parameters: the configuration of the neural network layers and its neuron weights. The experiment compares the best MLPs produced by the SA and FastSA using the accuracy and classifier complexity as comparison measures. The MLPs are optimized in order to produce a classifier for the MNIST database. The experiment showed that FastSA has produced a better MLP, using less computational time and less fitness evaluations.