Multi-objective Optimisation Of Rolling Rod Product Design Using Meta-modelling Approach
GECCO '02 Proceedings of the Genetic and Evolutionary Computation Conference
Hi-index | 0.00 |
Fuzzy fitness evaluation within evolutionary algorithms is increasingly being used to manage the fitness associated with genetic individuals as defined by the membership function of the fuzzy set. In most of the approaches reported in the literature, the fitness assignment mechanism is dependent on the defuzzified domain value or the degree of membership of the consequent solution set. This suggests that the trade-off between optimal solutions values and the truth-values (membership function values) of the associated solutions is not fully explored. This can result to a deception problem for search algorithm in cases where fuzzy fitness is used for evaluation. This paper presents a novel approach to deal fuzzy fitness evaluation within multi-objective optimisation framework to address the deceptive problem associated with current fuzzy fitness techniques reported in the literature. The proposed approach is applied to a rod rolling shape optimisation problem for automatic generation of optimal rod shapes.