摘要
在多目标进化算法理论和应用研究进展的基础上,通过采用外部群体的精英保留策略并引入拥挤距离来保持群体多样性的策略,设计了一种基于随机生成权向量的加权和函数的多目标进化算法。用所提出的方法求解若干常用的测试函数,并与NAGS-II进行比较,结果表明了算法的有效性。
On the basis of recent theory and application development of MOEAs,a new MOEA which utilizing the aggregated function of randomly generated weight vectors is presented.This proposed method adopts an external population to perform elitism and preserve diversity in the population by using the idea of crowding distance.The effectiveness of the suggested algorithm is demonstrated by performing some numerical experiments on some commonly used benchmark problems problem instances and comparing its results against NSGA-Ⅱ.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第32期4-6,163,共4页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(70301005)
教育部南开-天津大学刘徽应用数学中心资助项目。
