Publicação

Combining adaptation and optimization in bio-inspired multi-agent manufacturing systems

Ver documento

Detalhes bibliográficos
Resumo:Global markets impose strong requirements to manufacturing domain in terms of flexibility, robustness and reconfigurability. The multi-agent systems (MAS) paradigm is suitable to handle such requirements, introducing an alternative way to design complex, agile and adaptive systems. However, MAS based solutions may suffer of myopia due to the local optimal decision-making performed by the autonomous distributed agents having a partial knowledge of the problem. This paper depicts the optimization problem in MAS, particularly having in mind the achievement of adaptation, and explores the contributions that biology can offer to handle this issue. Two bio-inspired MAS solutions for routing pallets in a real assembly system are described to illustrate how optimization and adaptation can be combined.
Autores principais:Barbosa, José
Outros Autores:Leitão, Paulo; Pereira, Ana I.
Assunto:Multi-agent systems Manufacturing Systems Bio-inspiration Self-organization
Ano:2011
País:Portugal
Tipo de documento:comunicação em conferência
Tipo de acesso:acesso restrito
Instituição associada:Instituto Politécnico de Bragança
Idioma:inglês
Origem:Biblioteca Digital do IPB
Descrição
Resumo:Global markets impose strong requirements to manufacturing domain in terms of flexibility, robustness and reconfigurability. The multi-agent systems (MAS) paradigm is suitable to handle such requirements, introducing an alternative way to design complex, agile and adaptive systems. However, MAS based solutions may suffer of myopia due to the local optimal decision-making performed by the autonomous distributed agents having a partial knowledge of the problem. This paper depicts the optimization problem in MAS, particularly having in mind the achievement of adaptation, and explores the contributions that biology can offer to handle this issue. Two bio-inspired MAS solutions for routing pallets in a real assembly system are described to illustrate how optimization and adaptation can be combined.