| 求解复杂多目标优化问题MOEA/D-GEP算法 |
| |
| 引用本文： | 张冬梅,龚小胜,戴光明,彭雷.求解复杂多目标优化问题MOEA/D-GEP算法[J].华中科技大学学报(自然科学版),2012,40(4):33-36. |
| |
| 作者姓名： | 张冬梅 龚小胜 戴光明 彭雷 |
| |
| 作者单位： | 1. 中国地质大学武汉计算机学院,湖北武汉,430074
2. 中国地质大学武汉计算机学院,湖北武汉430074 华中科技大学计算机学院,湖北武汉430074 |
| |
| 基金项目： | 国家自然科学基金资助项目，湖北省自然科学基金资助项目，中央高校基本科研业务费专项资金资助项目 |
| |
| 摘 要： | 针对复杂多目标优化问题,提出一种基于演化建模的MOEA/D（基于分解的多目标遗传算法）求解算法（MOEA/D-GEP）.该算法利用MOEA/D算法思想分解多目标优化问题,对分解后得到的可行解用基于模拟退火的GEP算法建模,从中选取预测值较好的点进入下一次真实适应值的计算.采用国际公认的ZDT,DTLZ等测试函数进行实验验证,并与MOEA/D-EGO演化多目标优化算法进行了比较.实验结果表明：该算法在IGD性能指标上有较好的表现,说明将演化建模技术引入MOEA/D算法提高了种群个体分布模型的精度,降低了求解复杂多目标优化问题的计算成本.
|
| 关 键 词： | 复杂多目标优化问题 全局优化算法 基于表达式编程 演化多目标优化 MOEA/D-GEP |
MOEA/D-GEP algorithm solving expensive multiobjective optimization problems |
| |
| Zhang Dongmei,Gong Xiaosheng,Dai Guangming,Peng Lei.MOEA/D-GEP algorithm solving expensive multiobjective optimization problems[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2012,40(4):33-36. |
| |
| Authors: | Zhang Dongmei Gong Xiaosheng Dai Guangming Peng Lei |
| |
| Affiliation: | 1,2(1 School of Computer,China University of Geosciences,Wuhan 430074,China;2 School ofCcomputer,Huazhong University of Science and Technology,Wuhan 430074,China) |
| |
| Abstract: | An algorithm of GEP(MOEA/D-GEP) based on MOEA/D was proposed to solve the expensive problems,in which MOEA/D was used to decompose the multi-objective into a number of single optimization subproblems.The model for solutions with GEP was built.Then the better solutions were selected for the next optimization.This algorithm adopted internationally recognized testing functions to verify experiments,such as ZDT and DTLZ.The experiment results show that MOEA/D-EGO is good in solving expensive multi-objective problems and provides a good idea to solve expensive multi-objective optimization problems. |
| |
| Keywords: | expensive multi-objective optimization problems efficient globas optimaizatipn(EGO) gene expression programming(GEP) multiobjective optimization evolutionary algorithm based on decompostion(MOEA/D)GEP |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
