摘要
基于生物免疫系统克隆选择机理和独特型免疫网络理论,提出了一种新的免疫算法——克隆选择调节算法(CSAA).其主要特点是在克隆选择算法的基础上,引入了抗体的促进与抑制动态调节思想.通过运用自适应柯西变异、免疫记忆和克隆抑制等机制,该算法更好地保持了种群的多样性,提高了全局收敛的速度,从而有效避免了早熟现象.本文利用随机过程理论作为数学工具,采用纯概率方法证明了CSAA的概率弱收敛性.对该算法与其他克隆选择算法进行了仿真比较实验;仿真结果不仅验证了CSAA理论上的概率弱收敛性结论,同时也表明了该算法在求解多模态函数优化问题时具有更好的收敛性能和稳定性,更为有效可行.
Based on the elonal selection principle and idiotypie immune network theory exhibited in biological immune system, a new immune algorithm-the elonal selection adjustment algorithm （CSAA） is proposed. The algorithm introduces the dynamic adjustment idea of antibody promotion and suppression based on the elonal selection algorithm. With the use of the serf-adaptive Cauehi mutation operation, immune memory and elonal restrain mechanisms, CSAA holds a better diversity of population, improves the global convergence speed, and can effectively avoid prematurity. Stochastic processes theory is used as mathematical tools, and the weak convergence in the probability of CSAA is proved on the basis of pure probability method. The contrasting simulation experiments between CSAA and other clonal selection algorithms are performed. The experiment results verify that the theory proven above is right, indicate that CSAA has a preferable convergence ability and stability in solving multi-modal function optimization, and is more effective and feasible.
出处
《信息与控制》
CSCD
北大核心
2007年第4期476-480,485,共6页
Information and Control
基金
北京市教委重点学科共建资助项目
关键词
克隆选择算法
多模态函数优化
多样性
clonal selection algorithm
multi-modal function optimization
diversity
