自适应克隆免疫算法应用于宽带匹配网络设计

Application of Adaptive Clonal Immune Algorithm in Broadband Matching Networks

针对克隆选择算法的不足,提出了一种自适应克隆免疫算法.直接根据抗体的多样性来确定克隆规模,通过引入抗体的自适应变异操作,增加对精英抗体的记忆和相似抑制操作来产生种群新个体,以保证算法的全局收敛性和抗体的多样性.将该算法应用于宽带匹配网络的拓扑结构和元件值的综合优化设计中,并分析了优化过程中频率采样点数的合理选取问题.仿真结果通过与实频法、遗传算法和克隆选择算法相比较,表明该算法具有更快的收敛速度,设计的匹配网络具有更优的匹配性能.

Due to the shortage of clonal selection algorithm,an adaptive clonal immune algorithm（ACIA） was presented.The number of clones was directly determined by the diversity of antibodies.To generate population new individual,the adaptive mutation operation and integrated resemble restrain strategy with the memory for elites operation were adopted to ensure the global convergence of the algorithm and the diversity of the antibodies.In the process of optimizing the topology and element parameters of broadband matching network,the amount of sampled frequency in the optimization was analyzed.By comparison with the real frequency method,genetic algorithm and clonal selection algorithm, simulation results demonstrate that the proposed algorithm shows better performance and faster convergence speed.