基于免疫网络调节机理的混沌超变异免疫算法  

Chaotic Hypermutation Immune Algorithms Based on Immune Network Regulatory Mechanism

作　　者：何宏[1] 钱锋[2] 

机构地区：[1]上海师范大学机械与电子工程学院,上海201418 [2]华东理工大学化学工程联合国家重点实验室,上海200237

出　　处：《系统仿真学报》2008年第11期2810-2813,2820,共5页Journal of System Simulation

基　　金：国家973计划(2002CB3122000);上海市教委项目(06DZ030);上海师范大学一般科研项目(SK200739)的资助

摘　　要：根据生物免疫系统中存在的免疫网络调节机理,提出了一种实数编码的混沌超变异免疫算法,该算法结合克隆选择原理和混沌理论建立新的混沌超变异操作,增强了算法局部搜索能力。同时基于免疫网络数学模型设计抗体的激励水平,并以此作为抗体群免疫网络调节的依据,保持了抗体群的多样性。最后将其应用于函数优化问题,结果表明该算法的收敛性能优于克隆选择算法,而且能够有效克服早收敛问题。On the base of the immune network regulatory mechanism in the biological systems, a chaotic hypermutation immune algorithm （CHIA） was proposed. Combining clonal selection principle with chaotic theory, a novel chaotic hypermutation operation was developed in CHIA to further strengthen the searching ability of the algorithm in the solution domain. Furthermore, the stimulation level of each antibody based on mathematic model of the immune network was devised and used as the condition for antibody population regulation to maintain the population adversity. Finally, CHIA was applied to solve function optimizationn problems. The simulation results show CHIA not only has better performance than clonal selection algorithm （CLONALG）, but can efficiently avoid premature convergence.

关 键 词：免疫网络 混沌 超变异 激励水平 抗体群多样性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]