分布式计算在天线阵方向图综合中的应用

Application of Distributed Computing in Radiation Pattern Synthesis for Array Antennas

采用矩量法(MoM)分析天线阵列的辐射方向图,并使用遗传算法对天线阵列的方向图进行综合。由于遗传算法具有内在并行性,利用MATLAB分布式计算服务器和并行计算工具箱进行分布式并行程序设计,实现了主从式并行遗传算法,形成了天线阵方向图仿真综合软件,加快了天线阵列方向图的综合速度,得到了令人满意的结果。与传统串行算法相比,基于分布式计算的主从式并行遗传算法具有收敛速度快、仿真效率高的特点。

An array antennas radiation pattern synthesis software package using genetic algorithm （GA） coupled with the moment methods （MoM） was developed. Because genetic algorithms had an inherent property of parallel processing, distributed parallel programs were designed to accelerate the running process. The master-slave parallel genetic algorithm was achieved using MATLAB Distributed Computing Server software and Parallel Computing Toolbox software. Compared to the serial genetic algorithm, the optimization technique based on parallel GA and MoM is an effective, accurate, high performance method.