多映射混沌免疫算法及其在天线阵方向图综合中的应用

A multi-mapping chaotic immune algorithm and its application in pattern synthesis for antenna arrays

针对基本免疫算法中产生初始抗体群的盲目随机性和冗余性、补充自然消亡抗体细胞的单一性以及现有搜索方式易陷入小区域局部收敛的问题,融合免疫算法和混沌优化算法各自的优点,提出了一种新的混合优化算法.该算法采用Hénon序列来生成抗体群,采用Logistic序列来产生并补充自然消亡部分的抗体细胞,2种不同规则的混沌序列使得抗体群具有足够的多样性,扩大了搜索范围;同时采用了Logistic映射混沌变异和Gauss变异相结合的混合变异,提高了算法的搜索效率和收敛速度,克服了早熟现象.阵列天线方向图综合是智能天线的一项重要技术,采用提出的算法对阵列天线方向图进行了综合,仿真结果表明,与现有算法相比,该算法优化能力强,能有效避免局部收敛并且收敛速度快.

The immune algorithm has several defects： the generation of the initial antibody community is blind, random and redundant; the addition of naturally vanished antibody cells is monotonous; finally, existing searching modes easily to fall into local convergence. This paper proposed a mixed optimal algorithm by fusing the advantages of both the immune algorithm and the chaotic optimal algorithm. In this method, the antibody community is produced using the Hénon sequence, but antibody cells that replace the naturally vanished ones are generated using the Logistic sequence. Two chaotic sequences produced in different ways bring adequate diversity to the antibody community. As a result, the searching area is widened. Meanwhile, the simultaneous mapping of Logistic mutations and Gaussian mutations improves Searching efficiency and convergence rates, thereby eliminating premature convergence. The pattern synthesis tech- nique for array antennas is one of the key technologies in smart antenna systems. The proposed algorithm is used in calculating the optimal pattern of an antenna array. Simulation results show that the proposed algorithm can remarkably improve optimization performance and avoids local convergence while producing a high convergence rate.