求解多目标优化问题的多智能体遗传算法

目的 在求解多目标优化问题时，总是希望获得尽可能多的Pareto解，且这些解能够较均匀地分布在目标空间的Pareto边界上。方法 通过引入智能体的概念，并将多个智能体组成的多智能体系统与经典遗传算法相结合，给出了一种求解多目标优化问题的多智能体遗传算法。结果 对每个智能体在其邻域内进行局部Pareto寻优操作，而不是在整个群体中进行Pareto寻优，从而保证了群体的多样性，并在一定程度上抑制了种群的早熟现象。结论 该方法能够找到问题的分布较均匀的Pareto最优解。

A Multi-Agent genetic algorithm for multi-objective programming

Aim It is desirable to find more Pareto-optimal solutions which scattered uniformly over Pareto-front in solving multiobjective programming.Methods The notion of Agent is introduced and an algorithm is proposed with the combination of multi-Agent system and genetic algorithm for multiobjective programming. Results In each generation, an individual is viewed as an unit with the ability of communicating, and the Pareto-optimal solutions are found in the neighbor of each individual instead of the whole population, which ensures its diversity and restrains the prematurity.Conclusion The proposed methods can find the more Pareto solutions for multiobjective problem with good uniformity.