论文部分内容阅读
通过在克隆选择过程中引入聚类竞争机制,提出了一种免疫聚类竞争的克隆选择算法。采用了抗体聚类、竞争扩增、克隆删除、体细胞高频变异、抗体循环补充等思想及相关算子操作,增强聚类族中的优秀个体获得克隆扩增实现亲和力成熟的机会,提高抗体群分布的多样性,在深度搜索和广度寻优之间取得了平衡。实验仿真及应用结果表明:该算法具有可靠的全局收敛性及较快的收敛速度,将其应用于冶金过程目标优化中取得较好的效果。
By introducing clustering competition mechanism in the process of clone selection, a clonal selection algorithm of immune cluster competition was proposed. Antibodies such as antibody clustering, competition amplification, cloning and deletion, high frequency variation of somatic cells, antibody cycle replenishment, and related operator operations were used to enhance the chances of elite individuals in the cluster being able to achieve affinity maturation by cloning and amplification, The diversity of group distribution balances the depth of search and breadth optimization. Experimental simulation and application results show that the proposed algorithm has a reliable global convergence and a fast convergence speed, and achieves good results when it is applied to the optimization of metallurgical process targets.