Fechar
Metadados

@MastersThesis{Barroca:2019:NeAdEv,
               author = "Barroca, Eric Demetrius de Castro",
                title = "A new adaptive evolutionary algorithm for design optimization",
               school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
                 year = "2019",
              address = "S{\~a}o Jos{\'e} dos Campos",
                month = "2019-05-20",
             keywords = "Generalized extremal optimization, adaptive evolutionary 
                         algorithms, design optimization, space engineering, 
                         multidisciplinary optimization, otimiza{\c{c}}{\~a}o extrema 
                         generalizada, algoritmos evolutivos adaptativos, 
                         otimiza{\c{c}}{\~a}o de projetos, engenharia espacial, 
                         otimiza{\c{c}}{\~a}o multidisciplinar.",
             abstract = "In this work a new adaptive evolutionary algorithm derived from a 
                         stochastic algorithm for design optimization called Generalized 
                         Extremal Optimization (GEO) is introduced. It eliminates the 
                         single free parameter of GEO by controlling its value during the 
                         search by an adaptive approach which improved GEO performance 
                         significantly, even when considering the best GEO configurations. 
                         Nonetheless, it maintains the algorithm principal characteristics 
                         of dealing with continuous, discrete and integer design variables 
                         on convex or disjoint spaces while respecting design constrains. 
                         This new algorithm, called Adaptive Generalized Extremal 
                         Optimization (A-GEO), is implemented in two variations and applied 
                         to a multidisciplinary optimization problem of spacecraft 
                         engineering, showing the potential of the new methods in solving 
                         real engineering problems. RESUMO: Neste trabalho um novo 
                         algoritmo evolutivo adaptativo derivado de um algoritmo 
                         estoc{\'a}stico para otimiza{\c{c}}{\~a}o de projetos chamado 
                         Generalized Extremal Optimization (GEO) {\'e} introduzido. Este 
                         elimina o {\'u}nico par{\^a}metro livre presente no GEO 
                         atrav{\'e}s de um m{\'e}todo adaptativo que controla os valores 
                         deste durante a busca, assim melhorando a performance do GEO 
                         significantemente, at{\'e} mesmo quando comparada a sua melhor 
                         configura{\c{c}}{\~a}o. Por{\'e}m, mant{\'e}m suas principais 
                         caracter{\'{\i}}sticas de lidar com vari{\'a}veis de projeto 
                         continuas, discretas e inteiras em espa{\c{c}}os convexos ou 
                         disjuntos respeitando as restri{\c{c}}{\~o}es de projeto. Este 
                         novo algoritmo, chamado Adaptive Generalized Extremal Optimization 
                         (A-GEO), {\'e} implementado em duas varia{\c{c}}{\~o}es e 
                         aplicado a um problema de otimiza{\c{c}}{\~a}o multidisciplinar 
                         de engenharia de sat{\'e}lites, mostrando o potencial dos novos 
                         m{\'e}todos em resolver problemas reais de engenharia.",
            committee = "Santos, Walter Abrah{\~a}o dos (presidente) and Sousa, Fabiano 
                         Luis de (orientador) and Chagas, Ronan Arraes Jardim (orientador) 
                         and Ramos, Fernando Manuel (orientador) and Galski, Roberto Luiz 
                         and Chaves, Antonio Augusto",
         englishtitle = "Um novo algor{\'{\i}}tmo evolut{\'{\i}}vo adaptivo para 
                         otmiza{\c{c}}{\~a}o de projetos",
             language = "en",
                pages = "103",
                  ibi = "8JMKD3MGP3W34R/3TAJS9H",
                  url = "http://urlib.net/rep/8JMKD3MGP3W34R/3TAJS9H",
           targetfile = "publicacao.pdf",
        urlaccessdate = "29 nov. 2020"
}


Fechar