基于变异操作的蚁群算法用于连续函数优化  

Ant colony algorithm with mutation operation for continuous function optimization

作　　者：高芳[1,2] 韩璞[1] 翟永杰[1] 

机构地区：[1]华北电力大学控制与计算机工程学院,河北保定071003 [2]河北大学电子信息工程学院,河北保定071002

出　　处：《计算机工程与应用》2011年第4期5-8,共4页Computer Engineering and Applications

基　　金：河北省自然科学基金(No.F2009000215);华北电力大学校内基金项目(No.200814003)~~

摘　　要：介绍了基本蚁群算法的数学模型,在一种新的连续空间分解方法的基础上,对信息素更新方式进行了改进,采用信息素局部更新和自适应的信息素全局更新相结合的方式,以提高算法的收敛速度。引入了进化算法中的变异操作,对寻优过程中每次迭代的最优解进行变异,增加了种群的多样性,避免算法的早熟,以提高改进后蚁群算法的全局收敛性能。实验结果表明,提出的基于变异操作的蚁群算法在连续函数寻优上有更好的收敛速度和收敛性能。The mathematical model of basic ant colony algorithm is introduced.Based on a new allocation of cities,the pheromone updating rules are improved.The local pheromone updating rule and the adaptive global pheromone updating rule are combined so that the convergence rate is improved.In order to enhance the global convergence performance of the improved ant colony algorithm and avoid the precocious result,the mutation is introduced.Once the optimal solution of each iteration is gained,the mutation operation is applied to the optimal solution so that the population varieties are increased.The numerical simulation results demonstrate that the ant colony algorithm with mutation operation has faster convergence rate and better convergence performance for continuous space optimization problems.

关 键 词：蚁群算法 连续函数优化 自适应 变异 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]