遗传算法对多层旋波媒质本构关系的优化

运用遗传算法对连续变化的多层旋波媒质的本构关系进行了优化,设计了新的编码方案,用波分法对与编码相应的多层旋波媒质的电磁散射特性进行计算,通过进化筛选,得到了由多层旋波媒质材料构成的具有较好吸波性能的本构关系曲线。结果表明遗传算法在优化设计中的编码灵活,具有全局搜索等优越性.

Optimizing the Construction Relations of Chiral Slab via Genetic Algorithm

A novel method of optimizing the continuous variable constitutive relation of multilayered chiral slab based on genetic algorithm is discussed in this paper. The method derive automatically the continuous variable constitutive relations. The reflection coefficient is calculated by wave-splitting method. Further more, the genetic algorithm is used to search the maximum absorbability of the chiral slab with certain thickness. The one point crossover GA and two points crossover GA have both been tested in the research, the latter one is really more stable than the first one and its very important to set the proper crossover ration and mutation ration. The advantage of GA would be presented more obviously in designing artificial chiral material model with more variables.