Projeto de sistemas de engenharia utilizando o algoritmo de Busca Fractal Estocástica

• Ano de defesa: 2021
• Autor(a) principal: Bernardes, Renata
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Engenharia Mecânica
• Palavras-chave em Português: Projeto de sistemas de engenharia; Busca Fractal Estocástica; Otimização; Parâmetros Auto-Adaptativos; Modelos Caóticos de Busca; CNPQ::ENGENHARIAS::ENGENHARIA MECANICA::PROJETOS DE MAQUINAS::METODOS DE SINTESE E OTIMIZACAO APLICADOS AO PROJETO MECANICO
• Link de acesso: https://repositorio.ufu.br/handle/123456789/32674; http://doi.org/10.14393/ufu.di.2021.5568

Engineering systems design has an important role in development of world as we know it today. In this context, the optimization is configured as a set of tools that help the decision making during the design stage. In this scenario, the study of Heuristic Methods is essential for new and versatile tools be developed. In this contribution the aims is to study and apply the Stochastic Fractal Search (SFS) algorithm, which is based on fractal theory, to engineering systems design. Furthermore, operators to increase population diversity, to prevent premature convergence and to reduce computational cost are proposed. For this purpose, the association between the BFE algorithm with Search Chaotic Models and a strategy to reduce population size along generations was proposed. To evaluate the quality of obtained results, the sensibility analysis of the algorithm parameters, its application in single and multi-objective case studies, as well as the updating of its parameters were performed. The results obtained demonstrate the potential of this recent evolutionary approach, as well as the proposed operators to engineering systems design.