| 基于降维扫描方法的自适应多目标遗传算法 |
| |
| 引用本文： | 王晓兰,田宏亮,王慧中,杨琳琳.基于降维扫描方法的自适应多目标遗传算法[J].计算机工程与应用,2007,43(18):78-81. |
| |
| 作者姓名： | 王晓兰 田宏亮 王慧中 杨琳琳 |
| |
| 作者单位： | 兰州理工大学电气工程与信息工程学院 兰州730050 |
| |
| 摘 要： | 为了有效地应用遗传算法解决H2/H∞鲁棒控制系统设计问题,将遗传算法与局部优化方法相结合,提出了基于降维扫描方法的自适应多目标遗传算法(DRSA-MOGA)。通过引入适应度函数标准化方法、基于最优Pareto解集搜索的降维扫描方法和适应度函数自适应调整方法,提高了算法的全局优化性能和局部搜索能力。仿真结果表明,DRSA-MOGA算法在不损失解的均匀度的情况下可以达到很高的逼近度。
|
| 关 键 词： | 多目标优化 遗传算法 H2/H∞控制 局部搜索 Pareto最优解 |
| 文章编号： | 1002-8331（2007）18-0078-04 |
| 修稿时间： | 2006-11 |
Adaptive multi-objective genetic algorithm based on dimension reduction and scanning approach |
| |
| WANG Xiao-lan,TIAN Hong-liang,WANG Hui-zhong,YANG Lin-lin.Adaptive multi-objective genetic algorithm based on dimension reduction and scanning approach[J].Computer Engineering and Applications,2007,43(18):78-81. |
| |
| Authors: | WANG Xiao-lan TIAN Hong-liang WANG Hui-zhong YANG Lin-lin |
| |
| Affiliation: | College of Electrical Engineering and Information Engineering,Lanzhou University of Technology,Lanzhou 730050,China |
| |
| Abstract: | Combining genetic algorithm with local search,adaptive Multi-Objective Genetic Algorithm based on the Dimension Reduction and Scanning Approach（DRSA-MOGA） is introduced in order to make multi-objective robust control system smoothly reach H2/H∞ objective optimal solutions. By introducing three methods that are the method of fitness function normalization,the dimension reduction and scanning method and the adaptive adjust method of fitness function,the performance of global optimization and the ability of local search are improved. Simulation results show that Pareto solutions obtained by applying DRSA-MOGA can achieve a very high approximation,and can meanwhile keep good diversity. |
| |
| Keywords: | multi-objective optimization genetic algorithm H2/H∞ control local search Pareto optimal solution
|
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机工程与应用》浏览原始摘要信息 |
|
点击此处可从《计算机工程与应用》下载全文 |
