基于最佳进化方向的多目标遗传算法

该文模拟自然界中生物总是向着有利于自己的方向进化,即朝生物利益最大化的方向进化这一现象,给出了一种新的设计适应度函数的方法,并且结合多目标优化的Pareto最优解的概念,提出了求解多目标优化问题的一种新的算法———基于最佳基因的多目标遗传算法。数值实验表明,该算法不仅操作简单、鲁棒性强、速度快、且能够获得数量多而且广泛的Pareto最优解。

Multi-objective Genetic Algorithm Based on the Optimal Evolution Direction

In the nature,there is a phenomenon that biology always evolves toward the optimal direction,namely,which maximizes the total profits of biology.This paper gives a new strategy of fitness function combining the phenomenon with the Pareto optimal conception of multi-objective problem.A new genetic algorithm that solving the Pareto optimal set of multi-objective problem is presented.At last,the paper gives numerical test,and people can see that the proposed algorithm is simple,robust and fast,and it can get more and more extensive Pareto optimal set.