| 一种有效的基于实数编码的多目标遗传算法 |
| |
| 引用本文： | 田小梅,郑金华. 一种有效的基于实数编码的多目标遗传算法[J]. 湘潭大学自然科学学报, 2005, 27(2): 70-76 |
| |
| 作者姓名： | 田小梅 郑金华 |
| |
| 作者单位： | 1. 湖南环境生物职业技术学院,衡阳,421005
2. 湘潭大学信息工程学院,湖南,湘潭,411105 |
| |
| 基金项目： | 国家自然科学基金资助项目(90104021) |
| |
| 摘 要： | 针对实数编码遗传算法提出了一种通用的基于决策变量的复合交叉算子,并将之用于多目标优化问题的求解,算法效果良好,一定程度上解决了高维多目标优化问题在用遗传算法求解时收敛性差这一难题.通过实验首次揭示了交叉点数对多目标遗传算法性能的影响.
|
| 关 键 词： | 复合交叉 多目标遗传算法 多目标优化问题 基于变量的交叉 |
| 文章编号： | 1000-5900(2005)02-0070-07 |
| 修稿时间： | 2004-08-25 |
An Efficient Real-coded Multi-objective Optimization Genetic Algorithm |
| |
| TIAN Xiao-mei,ZHENG Jin-hua. An Efficient Real-coded Multi-objective Optimization Genetic Algorithm[J]. Natural Science Journal of Xiangtan University, 2005, 27(2): 70-76 |
| |
| Authors: | TIAN Xiao-mei ZHENG Jin-hua |
| |
| Affiliation: | TIAN Xiaomei1,ZHENG Jinhua2 |
| |
| Abstract: | This paper proposes a general compound-crossover operator which acts on decision variables in Real-Coded Genetic Algorithms. The algorithms which employ this crossover operator behave very well when they are used to solve Multi-objective Optimization Problem and , to a certain extent, solve the puzzle, i.e. Genetic Algorithms' bad convergence metric in dealing with high-dimension MOP. For the first time, we demonstrate that crossover-point count is closely related to the performance of Multi-objective Genetic Algorithms. |
| |
| Keywords: | compound crossover Multi-objective Genetic Algorithm Multi-objective Optimization variable-wise crossover |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
