TEDE Communidade:
http://www.bdtd.uerj.br/handle/1/3700
2024-03-29T06:56:29ZDesenvolvimento de ferramentas para modelar a dispersão de contaminantes atmosféricos com dados públicos brasileiros, usando o modelo AERMOD
http://www.bdtd.uerj.br/handle/1/21095
Título: Desenvolvimento de ferramentas para modelar a dispersão de contaminantes atmosféricos com dados públicos brasileiros, usando o modelo AERMOD
Autor: Guzmán, Nely Grillo
Primeiro orientador: Assis, Joaquim Teixeira de
Abstract: The relevance of simulating the dispersion of atmospheric contaminants is widely accepted by the scientific community and society. AERMOD is a contaminant dispersion model that needs to run some preprocessors that prepare the necessary input data. The information available internationally to run these preprocessors has resolutions of 1000m, which makes simulations difficult in small regions (a few kilometers around the source of emissions) and with irregular relief. Brazilian public data have higher resolution (30 m), but with them it is not possible to run the AERSURFACE and AERMAP preprocessors, verifying the need to create tools that could replace them. DATASURFACE and DATAMAP were created, which manage to generate the necessary data for the execution of AERMOD using publicly available data in Brazil. The results show that DATASURFACE offers greater practicality for users who wish to use AERMOD and can use data from the surface characteristics of the study region, if available. Using the developed tools, three case studies were performed, two of them in regions with irregular relief and one example in a region with predominantly flat relief. In the third example, the results were similar to those obtained using AERMOD's own preprocessors. For the first two examples, with the use of DATAMAP, patterns of contaminants dispersion were obtained, in smaller regions, which respect the orography of the terrain and consider the effect of small elevations present near the source of
contaminants. This way was demonstrated the need to use the tools developed for this type of cases. The developed tools are very necessary if the study area is a few m2 , but they can also be used for larger regions with 1600 km2 , where similar results are obtained in both simulations.
Instituição: Universidade do Estado do Rio de Janeiro
Tipo do documento: Tese2023-03-28T00:00:00ZIdentificação de parâmetros e funções via abordagem bayesiana para detecção de melanomas
http://www.bdtd.uerj.br/handle/1/21064
Título: Identificação de parâmetros e funções via abordagem bayesiana para detecção de melanomas
Autor: Nogueira, Adson Rodrigo Braga
Primeiro orientador: Abreu, Luiz Alberto da Silva
Abstract: This work focuses on solving inverse problems in heat transfer, exploring their
application in estimating parameters and functions for detecting melanoma skin tumors.
Grounded in Bayesian methodologies, specifically Markov Chain Monte Carlo (MCMC),
which has played a crucial role in advancing engineering, it opens new possibilities for
tumor identification and treatment using the classical heat biotransfer equation. Heat
transfer in living tissues, particularly in tumor detection, poses a multidisciplinary challenge. The classical Pennes model stands out, allowing for the estimation of distinctive
thermal properties between healthy and unhealthy cells. However, the computational
implementation of these models faces challenges due to their complexity, resulting in considerable computational costs. In response, this dissertation proposes a two-dimensional inverse problem for the thermal characterization of melanoma tumors. The focus is on
considering realistic geometries and applying methodologies aimed at reducing computa tional costs. The use of Bayesian approaches is explored for their effective handling of uncertainties, providing a robust perspective for analysis. In conclusion, the relevance and consistency of the adopted approaches are highlighted, demonstrating the convergence of results obtained through different computational techniques, such as Wolfram Mathematica and COMSOL Multiphysics. The MCMC method proved consistent throughout the dissertation, successfully detecting tumor regions and providing information about the affected tissue’s geometry. Mesh refinement and uncertainty analysis proved essential for obtaining reliable estimates. The resulting distributions, mainly Gaussian and Gamma, offered valuable insights into parameters in healthy and tumor domains.
Instituição: Universidade do Estado do Rio de Janeiro
Tipo do documento: Dissertação2023-12-26T00:00:00ZSoluções híbridas de difusão e convecção-difusão com transformações integrais e diferenças finitas
http://www.bdtd.uerj.br/handle/1/20671
Título: Soluções híbridas de difusão e convecção-difusão com transformações integrais e diferenças finitas
Autor: Mascouto, Fabricio da Silva
Primeiro orientador: Knupp, Diego Campos
Abstract: The present article addresses several hybrid solutions (numerical-analytical) for
conjugate diffusion-convection heat transfer problems and conduction. Four heat transfer
cases with internal flow are discussed, going through different aspects, ranging from smooth,
corrugated and sinuous channel flows, as well as a microchannel problem taking into account
slip flow and temperature jump effects in the solid-liquid interface, which are typically
despised at the macroscale. Still considering the internal flow in smooth channel, specifically
in the eigenvalue problem, a problem that incorporates convective effects was also addressed,
in addition to the classic formulation of Sturm Liouville. A final case study involves a heat
conduction problem with an irregular geometry. A fictitious domain encompassing the entire
original domain is created to avoid explicit treatment. The conjugated and conduction
problems are solved using the Generalized Integral Transform Technique (GITT), with a
single-domain formulation. This formulation could be used due to the approach of the
coefficient with spatial variations introduced in the energy equation, thus allowing the non explicit treatment of the wall-fluid interaction regions, enabling the treatment of complex
geometries. The auxiliary eigenvalue problem, generated from the application of integral
transformations, is solved through the Finite Differences Method, while the generated
ordinary differential equation system is solved analytically, thus making it a hybrid technique.
The solutions obtained by GITT are compared throughout the work with results from the
COMSOL Multiphysics® platform. Thus, the hybrid methodology was able to solve the
proposed problems correctly, taking the COMSOL solutions as a reference. In addition, the
numerical solutions of the eigenvalue problems proved to be advantageous in relation to the
solutions with integral transformations already presented in the literature.
Instituição: Universidade do Estado do Rio de Janeiro
Tipo do documento: Tese2023-08-23T00:00:00ZApplication of computational intelligence techniques in optimization and inverse heat transfer problems
http://www.bdtd.uerj.br/handle/1/20495
Título: Application of computational intelligence techniques in optimization and inverse heat transfer problems
Autor: Jardim, Lucas Correia da Silva
Primeiro orientador: Silva Neto, Antônio José da
Abstract: Inteligência Computacional pode ser definida como uma coleção de métodos que
resolvem problemas por meio de inteligência, raciocínio e tolerância à imprecisão. Nesta
tese, técnicas de Inteligência Computacional, como metaheurísticas de otimização, redes neurais, lógica fuzzy e inferência bayesiana são utilizadas para resolver problemas
inversos de transferância de calor. Para isso, são apresentados cinco estudos de caso.
O primeiro estudo trata do problema de identificação de falhas de contato na união
de diferentes materiais. Isso é formulado como um Problema Inverso de Condução de
Calor e resolvido com uma Rede Neural Artificial. O segundo emprega o método Fuzzy
Levenberg-Marquardt com Regularização de Tikhonov para resolver um problema inverso
combinado de condição-radiação. No terceiro, o desempenho do método Otimização Global
Topográfica é comparado quando inicializado com dois geradores de pontos diferentes
e finalizado com dois métodos de busca local diferentes para as soluções de problemas
inversos de transferência de calor. O quarto estudo apresenta um método para prever o
valor ótimo do parâmetro de regularização de Tikhonov, resolvendo versões reduzidas dos
problemas inversos considerados, onde testes são realizados em três problemas inversos
de estimativa de função. No quinto estudo, é apresentada a formulação e solução de um
modelo de bio-transferência de calor do cérebro contendo um eletrodo de Estimulação
Profunda do Cérebro (DBS) e, além disso, essa solução é usada para resolver um problema
inverso de determinação das condutividades térmica e elétrica do tecido cerebral usando
medidas supostamente obtidas com um sensor dentro do DBS. Concluindo, esta tese
mostra as capacidades de diversas técnicas de Inteligência Computacional na resolução de
problemas inversos de transferência de calor. Os resultados apresentados contribuem para
duas principais áreas: o campo dos problemas inversos em transferência de calor, suas
aplicações em ciência e engenharia; e o desenvolvimento de procedimentos computacionais
eficientes baseados em técnicas de Inteligência Computacional.
Instituição: Universidade do Estado do Rio de Janeiro
Tipo do documento: Tese2023-07-05T00:00:00Z