Desenvolvimento de algoritmo para controle de tráfego urbano usando redes neurais e algoritmos genéticos

• Ano de defesa: 2009
• Autor(a) principal: Francisco, Marcus Vinícius Cardador
• Orientador(a): Giorno, Fernando Antonio de Castro
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Pontifícia Universidade Católica de São Paulo
• Programa de Pós-Graduação: Programa de Estudos Pós-Graduados em Tecnologias da Inteligência e Design Digital
• Departamento: Faculdade de Ciências Exatas e Tecnologia
• País: BR
• Palavras-chave em Português: Propagação reversa; Algoritmos; Genética; Tráfego; Algoritmos geneticos; Redes neurais (Computacao); Engenharia de trafego
• Palavras-chave em Inglês: Neural networks; Back-propagation; Algorithms; Genetic; Traffic
• Área do conhecimento CNPq: CNPQ::ENGENHARIAS
• Link de acesso: https://tede2.pucsp.br/handle/handle/18252

This research has as goal to introduce an alternative solution for vehicles traffic flow control. Researches on similar subjects around the world were taken as a basement for this study which makes use of a hybrid architecture. This architecture is composed by a back-propagation algorithm, which is responsible for creating and training the networks that will take care of traffic flow control, and a genetic algorithm, responsible for all chromosome relations which will generate new networks based on its previews parents. The results for this combined algorithms shows that errors were decreased if compared to the other researches described below. This makes this a plausible solution. The whole complexity involved on current study as well as on traffic flow control gives many possibilities for development of new solutions and improvements on traffic flow subject