基于Pareto的多目标优化问题是进化算法的一个重要研究方向,而如何构造Pareto非支配集则是提高算法效率的关键所在。通过对选举现象的观察,同时针对多目标个体之间的特性,提出了一种快速求解多目标Pareto非支配集的方法:选举法则(election principle,EP),分析了其时间复杂度为O(rmN),并对其进行了正确性证明。因为种群中实际的非支配个体数m比进化群体规模N小,所以与同类方法相比,EP有更高的效率,并通过了实验验证。
The multi-objective optimization problem based on pareto is a important research direction of the evolutionary algorithm,and how to improve the efficiency of constructing the Pareto non-dominated set is a key to the algorithm.This paper proposed a quick method of constructing multi-objective pareto non-dominated set through observing the election phenomenon and understanding the mutual character of multi-objective individual,namely the election principle(EP),analyzed that its computational complexity was O(...