摘要
维持群体多样性是提高进化算法性能的一个主要出发点。本文提出了一种基于免疫选择和自组织临界变异的进化算法。其中,利用免疫浓度调节设计的选择算子使算法在开发新解时能选到多样性的个体;基于自组织临界思想的变异算子使算法在探测新解时能在合理的模型指导下进行。针对几种典型的复杂函数优化问题的求解实验表明该算法在收敛速度和全局收敛性方面都较好。
Keeping populations diversity is the main point to improve performance of evolutionary algorithm. An evolutionary algorithm is presented on the basis of immune and self-organized criticality (SOC), which adopts the selection operator based on immune density adjustment in order to get different individuals and the mutation operator based on the idea of SOC to make its exploration under sound models. At last, experiments were given to solve several typical complicated function optimization problems. The results show that the algorithm has good performance in the aspects of both convergence speed and the global convergence.
出处
《系统仿真学报》
CAS
CSCD
2004年第8期1785-1788,共4页
Journal of System Simulation
基金
国家自然科学基金(60204009)。
关键词
进化算法
未成熟收敛
免疫
自组织临界
遗传操作
evolutionary algorithm
premature convergence
immune
self-organized criticality
genetic operation
