电路模拟吸收体的遗传算法设计

提出了一个应用遗传算法的电路模拟多层吸收体的新设计、遗传算法自动综合各层的电磁参数和厚度以及金属电路屏的各参数,并同时提供几个设计方案。当应用遗传算法对特宽带吸收体设计时,其迭代过程出现波动和在带内出现不好的结果。为了克服这些缺点,文中引入了两项技术,使其工作特性明显地改善。设计实例表明在无金属基底的吸收体情形,电路模拟吸收体比普通设计优越得多,在厚度更小和用更少材料下能在特宽带内获得低反射。

Design of Circuit Analog Absorber Using Genetic Algorithm*

This paper presents a new method for designing circuit analog (CA) multi-layer absorbers using modern genetic algorithm (GA). The GA automatically synthesizes the EM parameters and thickness of each layer and the parameters of the metal circuit screen provide several design schemes simultaneously. But when using the GA in designing extraordinary wide band absorbers, its iterative process will go to fluctuation and give some bad results within the band. For overcoming these drawbacks, two techniques are introduced to GA program, and the operating performance is improved obviously. The new method shows that it is of high design efficiency and of easy use. The design examples also show that in the absorber without metal base cases, the characteristics of CA absorbers are much superior than ordinary designs, which give low reflection within very wide band with a thinner thickness and less materials used.