基于距离收敛量和历史信息密度的多目标进化算法

在筛选个体的过程中,多目标进化算法大都利用非支配信息和密度信息评价个体。但当个体互为非支配关系时,上述信息就难以区分个体的优劣从而影响算法性能。为了改善上述情况,本文提出了一种基于距离收敛量和历史信息密度的多目标进化算法。距离收敛量可以在非支配信息不能区分个体时评价个体的收敛性；历史信息密度可以更精确的提供个体多样性信息。在与三个先进的多目标进化算法的对比实验中,新算法的求解质量明显优于对比算法。

A distance convergence and history density based multi-objective evolutionary algorithm

Most multi-objective evolutionary algorithms use Pareto dominant criterion and density criterion to select individuals from generation to generation. However, as the Pareto both criterions fail to discriminate the convergence and diversity degrees of individuals when the individuals are Pareto-optimal solutions. To address this issue, this paper proposes a distance convergence and history density based multi-objective evolutionary algorithm. The distance convergence criterion is defined to distinguish Pareto-optimal individuals. The history density criterion is present a more accurate density estimation. The experimental results show that our algorithm performs competitively with respect to chosen state-of-the-art designs.