一种用于多目标优化的克隆选择算法  

Clonal Selection Algorithm for Multi-objective Optimization

作　　者：李恒杰[1] 郝晓弘[1] 张磊[1] 

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050

出　　处：《科学技术与工程》2008年第2期387-391,397,共6页Science Technology and Engineering

基　　金：甘肃省科技攻关(GS044-A52-001-24);甘肃省自然科学基金项目(3ZS042-B25-012)资助

摘　　要：多目标进化算法的研究目标是使算法种群快速收敛并均匀分布于问题的Pareto最优前沿面。为此,在比较与分析已有多目标进化算法的基础上,借鉴免疫系统中的克隆选择原理,提出了一种用于多目标优化的克隆选择算法。该方法只对部分当前所得到的Pareto最优解进行进化操作,并选用一种简单的多样性保护机制来保证Pareto最优解具有良好的分布特征。实验结果表明该方法能够很好地达到Pareto最优前沿面,较好地保持解的多样性,并且具有很快的收敛速度。The evolutionary multi-objective optimization techniques are used to find the non-dominated set of solutions and distribute them uniformly in the Pareto optimal front. After comparing and analyzing the developing history of evolutionary multi-objective evolutionary algorithms, a multi-objective clonal selection algorithm is proposed, based on clonal selection principle of immune system. Only some Pareto optimal solutions are selected for further evolutionary operation, and a simple mechanism is used to maintain a good spread of Pareto optimal solutions in the algorithm. It is shown by experimental results that the method can reach the Pareto optimal front very well, retain the diversity of the solutions better, and use much less time.

关 键 词：多目标进化算法 克隆选择算法 多目标优化 PARETO最优解 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]