一种协同进化模型的研究  

Study on cooperation evolution model

作　　者：刘小勇[1] 刘军[1] 赵桂荣[1] 马静[1] 王勇超[2] 

机构地区：[1]西安理工大学研究生学院,西安710048 [2]西安理工大学计算机科学与工程学院,西安710048

出　　处：《计算机工程与应用》2012年第8期51-54,72,共5页Computer Engineering and Applications

摘　　要：在分析了目前的协同进化算法的优越性与不足的基础上,借鉴生命科学中协同进化的生态背景,提出了一种改进的协同进化算法模型。该模型的主导思想在于将一类物种的某种进化特性作用于另一类物种相似的进化特性,以提高该物种的进化过程。用双物种的聚类算法进行了简单的仿真实验,结果证明在该算法模型下,两物种的最优解优于各自单独进化的结果,且寻优时间大大缩短。On the basis of analyzing the present cooperative evolutionary algorithm superiority and disadvantage, with cooperative evolution of the ecological background in the life sciences, an improved model of co-evolutionary algorithm is proposed. The core of this model is that a species of the characteristics of a role in the evolution of other species similar to the evolution of properties, in order to enhance the evolution of the species. It carries out a simple simulation with two species of the clustering algorithm, the results show that in the algorithm model, the optimal solutions of the two species are better than the results of their separate evolutionary, the time of searching for the best result is improved a lot.

关 键 词：人工免疫 协同进化 生态环境 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]