Publicação

Multiobjective optimization of polymer extrusion: decision making and robustness

Ver documento

Detalhes bibliográficos
Resumo:A Multi-Objective Evolutionary Algorithm (MOEA) is used to optimize polymer single screw extrusion. In this approach, the MOEA is linked to a modelling routine that quantifies the objectives as a function of the decision variables (i.e., operating conditions and/or screw geometry). Due to the conflicting nature of some objectives, the optimization algorithm uses a set of possible solutions to the problem that evolves during suc-cessive generations to a set of optimal solutions denoted as Pareto set. Since practical process optimization should yield a single solution, it is convenient to implement also a Decision Making (DM) strategy. Two methodologies were followed. In one case, the solutions were selected based on the preferences of a decision maker. Alternatively, the sensitivity of the solutions to small changes in the design variables was taken into account through a robustness analysis. The analysis of various case studies and the comparison with experi-mental data validated the method and demonstrates its potential.
Autores principais:Gaspar-Cunha, A.
Outros Autores:Covas, J. A.; Denysiuk, Roman; Recio, Gustavo
Assunto:Optimization Extrusion
Ano:2016
País:Portugal
Tipo de documento:comunicação em conferência
Tipo de acesso:acesso aberto
Instituição associada:Universidade do Minho
Idioma:inglês
Origem:RepositóriUM - Universidade do Minho
Descrição
Resumo:A Multi-Objective Evolutionary Algorithm (MOEA) is used to optimize polymer single screw extrusion. In this approach, the MOEA is linked to a modelling routine that quantifies the objectives as a function of the decision variables (i.e., operating conditions and/or screw geometry). Due to the conflicting nature of some objectives, the optimization algorithm uses a set of possible solutions to the problem that evolves during suc-cessive generations to a set of optimal solutions denoted as Pareto set. Since practical process optimization should yield a single solution, it is convenient to implement also a Decision Making (DM) strategy. Two methodologies were followed. In one case, the solutions were selected based on the preferences of a decision maker. Alternatively, the sensitivity of the solutions to small changes in the design variables was taken into account through a robustness analysis. The analysis of various case studies and the comparison with experi-mental data validated the method and demonstrates its potential.