摘要
介绍了一种新的多目标进化算法——Pareto-MEC。将基本MEC和Pareto思想结合起来处理多目标问题。提出了局部Pareto最优解集与局部Pareto最优态集概念,并利用概率论的基本理论证明了趋同过程产生的序列强收敛于局部Pareto最优态集。数值试验验证了Pareto-MEC算法的有效性。
Pareto mind evolutionary computation (Pareto-MEC) is a new multi-objective evolutionary algorithm (MOEA), which introduces the theory of Pareto into MEC for multi-objective optimization. Feasibility and efficiency of Pareto-MEC are illustrated by numerical results. The concepts of local Pareto optimal solution set and local Pareto optimal state set are presented. And it is proved that the sequence of population generated through operation similartaxis strongly converges to local Pareto optimal state by using the probability theory.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第10期233-236,共4页
Computer Engineering
基金
国家自然科学基金资助项目(60174002)
北京市教育委员会科技发展计划基金资助项目(KM200600006001
KM200600006003)
关键词
进化算法
多目标优化
思维进化计算
收敛性
趋同操作
异化操作
Evolutionary computation
Multi-objective optimization
Mind evolutionary computation
Convergence
Operation similartaxis
Operation dissimilation