一种克服遗传算法早熟的参数调整及并行方法

Adjustment of Parameters and Parallel Reality of A Genetic Algorithm Based on Avoiding Premature Convergence

遗传算法是一种自适应全局优化概率算法 ,容易产生早熟 (过早收敛 )现象 ,影响了问题的求解 .本文试图借助于多种群进化和种群间个体移植的概念 ,通过自适应控制参数的调整 ,利用移植并行的方法求出问题的最优解 (或近似最优解 )以便避免早熟 ,从而提高算法的搜索范围和效率 .

Genetic algorithms is a kind of adaptive global optimization statistical algorithm, it is tend to emerge premature convergence, which affects the solving of problems. In this paper, we attempt to avail of adjustment of adaptive parameters and the conception of multiple groups and transplant in different groups to avoid premature convergence efficiently.