基于免疫网络理论的动态超变异免疫算法  

Dynamic Hypermutation Immune Algorithms Based on Immune Network Theory

作　　者：何宏[1] 钱锋[1] 

机构地区：[1]华东理工大学化学工程联合国家重点实验室

出　　处：《华东理工大学学报（自然科学版）》2007年第3期429-435,共7页Journal of East China University of Science and Technology

基　　金：国家杰出青年科学基金(60625302);国家973计划基金(2002CB3122000);国家863计划项目(20060104Z1081);上海市科委重大基础研究基金(05DJ14002);上海市教委科研项目(06DZ030)

摘　　要：基于免疫网络理论,提出了一种动态超变异免疫算法,该算法通过采用新的超变异方法增强了算法在解域的搜索能力。同时根据抗体的激励水平进行免疫调节操作,保持了抗体群的多样性。最后根据随机过程的理论知识,证明了该算法的收敛性。仿真结果表明:该算法采用格雷编码时的性能优于用二进制编码实现的算法,与遗传算法和克隆选择算法相比,不仅收敛速度快,而且全局搜索能力强。A new dynamic hypermutation immune algorithm （DHIA） is proposed on the basis of immune network theory. DHIA adopts novel hypermutation operation to further strengthen the searching ability of the algorithm in the solution domain. Immune regulatory operation based on the stimulation level of each antibody is also applied to effectively maintain the diversity of the population and avoid the premature convergence. Furthermore, the convergence of the DHIA is proved theoretically through the Markov stochastic process theory. The simulation experiments show DHIA with gray coding outperforms its traditional binary counterpart. It is also verified that DHIA has better convergent performance than clonal selection algorithm （CLONALG） and genetic algorithms with elite strategy for global optimization problems.

关 键 词：免疫网络 超变异 激励水平 收敛性 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]