基于小生境粒子群优化的E面宽带波导滤波器设计

Design of E- plane Broadband Waveguide Filter Based on Niching Particle Swarm Optimization

针对传统综合法和模式分析法宽带E面滤波器结构设计中,精度低,计算量大的问题,提出一种基于改进粒子群优化的宽带E面波导滤波器设计算法。该算法首先采用传统方法得到滤波器大概尺寸,并以此为中心设置粒子群算法搜索域,然后,基于S参数设计适应度函数,采用基于小生境粒子群优化的模式匹配法对结构尺寸寻优计算,并最终获得最优滤波的结构尺寸。仿真和实验表明,本文优化方法可快速获得宽带E面滤波器的尺寸,滤波器插损低且阻带抑制高。

The structure sizes of E- Plane broadband waveguide filter are designed usually based on the traditional synthesis or mode matching method. But the filter design either can hardly meet the precision needs or brings the heavy computational load. According to the problem,a design of E-Plane broadband waveguide filter based on niching particle swarm optimization is proposed. Firstly,the approximate size of the filter is gotten with the conventional synthesis and the size is set as the search field of the optimization. Then,using mode matching method accurately calculate the size of the S- parameters of different structures and the sizes of the structure are optimized with the improved niching particle swarm optimization to decrease the computational load,so as to quickly obtain the optimal filter sizes. Simulation and experiment show that the proposed optimization method can quickly get broadband E- plane filter sizes and the filter responses low insertion loss and high stop- band rejection.