摘要
针对田口算法(Taguchi method,TM)处理复杂优化问题时易早熟收敛的缺点,引入遗传算法中的变异思想,提出了一种基于变异机制的田口算法(mutation-based Taguchi method,MTM)。在基本田口算法的基础上,利用变异算子生成参数的水平值,提高算法跳出局部极值的能力。同时,采用自适应内循环机制,多样化算法搜索空间。将改进后的田口算法应用到多层吸波材料的优化设计中。实验结果表明,变异田口算法能够有效跳出局部极值,寻找到全局最优值,综合考虑各个变异方式的寻优效率、寻优精度及稳定性,参数的2水平值高斯变异性能最佳。优化后的多层吸波材料能够达到低反射系数、薄厚度的设计要求。
Aiming to overcome the disadvantage that premature convergence was easily appeared when Taguchi method (TM) was applied to handling complex optimal problems,the variation idea of Genetic algorithm was introduced here and a Mutation-based Taguchi method (MTM) was proposed.In order to improve the ability of the algorithm to jump out of local optima,on the basis of TM,using the mutation operators generate the level values of parameters.Meanwhile,adaptive inner loop mechanism was used to diversify the search space.Then the improved Taguchi method was applied to the multilayer absorber design and optimization.Simulation result indicates that MTM can effectively escape from the local extreme and find the global optimum value.Among all the mutation operators,2 level value Gauss mutation method has the best optimization efficiency,accuracy and stability optimization.Optimized multilayer absorbing material can achieve the goal of low reflectance and thin thickness.
出处
《科学技术与工程》
北大核心
2015年第1期250-256,共7页
Science Technology and Engineering
关键词
田口算法
变异
内循环
早熟收敛
多层吸波材料
Taguchi method
mutation
inner loop
premature convergence
multilayer absorber