Compensation Method In Genetic Algorithm For Multi-Objective Optimization

Yuan Hua,Chen Guo-Qing
2005-01-01
Abstract:Genetic algorithm (GA) in solving complex multi-objective optimization problems does not take sufficiently the effect of mathematical objective functions properties to chromosome evolution into consideration. Which makes fast convergence empower the objective function to affect much the fitness of chromosomes in evolution process. To such affection a compensation method is presented to renew the rank-based fitness evaluation GA. An example is showed and some interesting empirical results are obtained: more rational individuals survive and the set of pareto-solutions is enlarged.
What problem does this paper attempt to address?