FACILITY LAYOUT METHODS USING DISTRIBUTED EVOLUTION STRATEGIES

Evolution Strategies (ES) have more excellent property than the other existing meta-heuristics in a sense of easiness of coding and its manageability of self-adaptive strategy. Hirabayashi et al. proposed a facility layout method using ES and they recommended to solve a layout problem multiple times by changing the values of the recombination parameters for obtaining a good layout. In ES, it is know that the recombination parameters affect the performance of the ES quite directly. This paper proposes a facility layout method using Distributed ES. In the method, ESs with different recombination parameter values are performed in parallel. Some numerical experiments made it clear that the proposed method provide solutions with the same accuracy as the previous method even though the computation time of the proposed method is about one-fourteenth of the previous method.