论文部分内容阅读
针对标准粒子群(Particle Swarm Optimization,PSO)算法应用于非对称副瓣形状方向图综合时,收敛速度慢和容易早熟的缺陷,提出了一种改进标准粒子群算法。该方法借助于修正Taylor综合法先得到连续口径分布,然后通过对其抽样得到粒子群初始化的基本值,对该基本值添加随机值得到PSO优化的初始粒子种群,将该种群用于PSO迭代时,采用“精英”选择思想,即用较好的粒子替代部分较差的粒子,直到满足停止条件。文中给出了运用该方法综合的两个实例,验证了其可行性,并通过多次重复试验,验证了该方法的高效性。
Aiming at the defect that the standard Particle Swarm Optimization (PSO) algorithm is applied to the asymmetric sidelobe shape pattern synthesis, the convergence speed is slow and premature, an improved standard PSO algorithm is proposed. The method first obtains the continuous caliber distribution by means of modified Taylor synthesis method, and then obtains the basic value of particle swarm initialization by sampling it, and adds PSO to the basic value to get the PSO-optimized initial particle swarm, which is used in PSO iteration , Use “elite ” choice of thinking, that is, use better particles to replace some of the poorer particles, until the stop conditions are met. In this paper, two examples of the method are given. The feasibility of the method is verified. The efficiency of the method is verified by repeated experiments.