【作者】 吴盛辉；

【导师】 葛德彪；

【作者基本信息】 西安电子科技大学 ， 无线电物理， 2007， 硕士

【摘要】 粒子群优化算法(PSO)是一种基于群体搜索的自适应进化计算技术。由于其能有效解决复杂优化任务,在过去几年中获得了飞速的发展,并在图像处理、多目标优化和电磁场领域得到了广泛应用。本文深入研究了粒子群优化算法,讨论了算法中控制参数对算法搜索性能的影响,将粒子群优化算法和传播矩阵法结合应用于介质板电磁参数的反演。针对PSO算法优化复杂多峰函数时会收敛于局部极值,计算结果误差较大等缺点,对标准算法进行改进,数值模拟结果和算例充分说明了改进算法有很大改进。为了达到吸波和屏蔽的效果,电磁吸波混凝土材料通常为电和磁均有耗的介质。对于电和磁均有耗的介质,导出了相应的透入深度、传播波长、反射和透射系数Fresnel公式,分析了阻抗匹配的概念。重点讨论了设计吸波材料的两个关键因素:衰减特性和匹配特性。利用粒子群优化算法作为优化方法,分别计算了吸波混凝土材料达到吸收和屏蔽效果时,两层吸波混凝土材料介质参数和厚度的搭配。同时,编制了相应的计算机优化辅助设计软件界面,软件不仅能实现两层吸波材料达到吸波和屏蔽要求时吸波材料的设计,还具有操作简洁,便于用户操作使用的优点。通常吸波材料都是掺杂介质,吸波材料的配置须得测量和提取掺杂介质的等效电磁参数,以实现优化设计的结果。电磁参数提取Nicolson-Ross-Weir(NRW)方法作为微波测量中的一部分,从二十世纪七十年代提出已经三十余年。本文详细推导了矩形波导中和自由空间中利用NRW方法提取介质参数的公式,总结了提取方法,讨论了两种情况下提取方法的不同。通过FDTD和传播矩阵法分别计算了矩形波导中和自由空间中介质散射参数,利用数值模拟的方法提取介质参数,分析了NRW方法提取介质参数时所出现的谐振问题。最后通过PSO和NRW两种方法来提取掺杂混凝土介质的等效参数,并与经验公式计算结果相比较,结果相符。更多还原

【Abstract】 Particle swarm optimization(PSO) algorithm is a self-organizing evolutionary computation technique based on the movement of swarms. Because it can resolve complex optimization tasks effectively, PSO algorithm is developing very fast in the past few years, and has been applied in image disposing, multi-objection optimization and electromagnetism fields. This paper study PSO algorithm, and discuss control parameters in algorithm how to influence the search performance. PSO algorithm and transmitting matrix algorithm are applied to the reconstruction of electromagnetism parameters of medium. Because of the disadvantages, such as low efficiency and poor accuracy when PSO algorithm optimizes complex and multi-extremum function, two modified particle swarm optimization algorithms are proposed. Numerical value results and examples prove that modified algorithms are superior to the standard algorithm.In order to achieve absorbing electromagnetic wave and shielding effectiveness, absorbing materials of concrete is lossy in electromagnetic. This paper educe computation formula of permeating depth, wavelength, reflection and transmission coefficient when electromagnetic wave propagates in the lossy medium, and analyses the conception about matching impedance. Two keys about the design of absorbing material are discussed: attenuation and matching characteristic. PSO algorithm is used to compute electromagnetic parameters and depth of two layered media, when concrete media can achieve absorbing electromagnetic wave and shielding effectiveness. A soft about the design of absorbing material is developed. This soft is easy to use to the user.Because absorbing material usually is composite medium, the configuration of absorbing material must measure and extract equivalent electromagnetic parameters of composite media, to achieve the result of optimum design. The Nicolson-Ross-Weir, NRW, of electromagnetic parameters extraction, which is a part of microwave measuring, has been thirty years since it was invented. The NRW formula of electromagnetic parameters reconstruction when the medium is in rectangle wave-guide and free space are educed. The method is summarized. The difference about two cases is discussed. Using FDTD and transmitting matrix algorithm, the scattering parameters of medium in rectangle wave-guide and free space are computed, and electromagnetic parameters are extracted. The resonance problem in NRW is analysed. At last, the effective electromagnetic parameters of the composite media are reconstructed by two methods: PSO algorithm and NRW. The results of two methods compared with experience method are accurate.更多还原