摘要
基于人类免疫系统的机理提出一种进化算法.简述了算法的基本原理与特点,定义了克隆、超变异、选择和记忆4种基本操作算子,给出了算法的主要步骤,并证明了算法能够以概率1收敛到全局最优点.用不同的测试函数进行仿真实验,结果表明该算法是有效的,能以较快的速度完成给定范围的搜索和优化任务.
A new evolutionary algorithm, which simulates the natural immune response process, is proposed based on the human immune systems mechanism. The main idea and characteristics of the algorithm are addressed. Four operators such as clone, hyper-mutation, selection and memory are defined. The main steps of the algorithm are presented. It is proved that the algorithm can converge to the global optimal value with probability 1. In experiment, the algorithm is used to optimize different functions for testing and the result shows that the algorithm is valid and can perform searching in a given range and carry out the optimization task in high rate.
出处
《控制与决策》
EI
CSCD
北大核心
2004年第2期219-222,共4页
Control and Decision
关键词
免疫系统
免疫响应
进化算法
优化
immune systems
immune response
evolutionary algorithm
optimization
