MOEA/D-GO for Fragmented Antenna Design
MOEA/D-GO for Fragmented Antenna Design
复制标题
DOI：
10.2528/pierm13071610
复制
发表时间：
2013
影响因子：
1
通讯作者：
D. Ding;G. Wang
中科院分区：
文献类型：
--
作者：
D. Ding;G. Wang
本文提出了一种基于分解和增强遗传算子的混合多目标进化算法MOEA/D-GO (multiobjective evolutionary algorithm Based Based and Enhanced Genetic Operators)，用于碎片型天线的设计。结合遗传算法二维染色体编码的特点，将MOEA/D处理多目标优化问题的能力和效率相结合。此外，还引入了增强的遗传算子来产生新的个体。对6个多目标0/1背包问题的数值计算结果表明，采用加权和分解方法的MOEA/D- GO优于原始的MOEA/D和MOEA/D- pr （MOEA/D结合Path- Relinking算子）。然后将其应用于cpw馈电单极天线的优化，以实现带陷波特性。数值和试验结果表明，MOEA/D-GO算法在解决碎片化天线多目标优化问题上具有广阔的应用前景。
In this paper, a hybrid multiobjective evolutionary algorithm, MOEA/D-GO (Multiobjective Evolutionary Algorithm Based on Decomposition combined with Enhanced Genetic Operators), is proposed for fragment-type antenna design. It combines the ability and e-ciency of MOEA/D to deal with multiobjective optimization problems with the speciflc character of two-dimensional chromosome coding of genetic algorithm. And enhanced genetic operators are also introduced to generate new individuals. Numerical results of a set of six multiobjective 0/1 knapsack problems show that MOEA/D- GO with weighted sum decomposition approach outperforms original MOEA/D and MOEA/D-PR (MOEA/D combined with Path- Relinking operator). Then it's applied to optimize a CPW-fed monopole antenna to achieve band-notch characteristic. Both numerical and test results show that MOEA/D-GO is promising for solving multiobjective optimization problems about fragmented antenna.