Heuristicas e metaheuristicas para otimização combinatoria multiobjetivo
AUTOR(ES)
Jose Elias Claudio Arroyo
DATA DE PUBLICAÇÃO
2002
RESUMO
This work presents contributions to the development of heuristic methods for multiobjective combinatorial problems. The goal of the proposed methods is to generate in a reazonable time a set of approximately Pareto-optimal solutions, allowing the decision maker to choose a solution of interest. The methods are tested on the flowshop scheduling problem and the knapsack problem. First, we develop a constructive heuristic to generate a set of dominant solutions for the multiobjective flowshop scheduling problem. In order to find sets of dominant solutions which are closer to the Pareto-optimal sets, we develop a local search heuristic and two metaheuristics. The first one is based on Genetic Algorithms which strongly use the concept of Pareto dominance and combine elitism, population diversity and local search strategies. Then, we propose a Tabu Search-based method which explores a set of solutions in parallel to find a variety of solutions distributed along the Pareto frontier. The performance of the proposed methods is tested in a number of instances for the multiobjective flowshop scheduling problem and the multiobjective knapsack problem.
ASSUNTO(S)
algoritmos geneticos heuristica otimização combinatoria
ACESSO AO ARTIGO
http://libdigi.unicamp.br/document/?code=vtls000256530Documentos Relacionados
- Inteligencia computacional na sintese de meta-heuristicas para otimização combinatoria e multimodal
- Método de geração de colunas e meta-heurísticas para alocação de tripulação
- Heurísticas e algoritmos evolutivos para formulações mono e multiobjetivo do problema do roteamento multicast
- Um algoritmo genetico para otimização multiobjetivo fuzzy
- Algoritmos determinístico e evolucionário intervalares para otimização robusta multi-objetivo