基于信息素扩散机制的双种群蚁群优化算法  

Dual Population Ant Colony Optimization Based on Pheromone Diffusion

作　　者：朱凯燕[1] 成諝虓[1] 

机构地区：[1]同济大学信息管理与信息系统系,上海201804

出　　处：《微计算机应用》2010年第7期22-26,共5页Microcomputer Applications

摘　　要：针对基本双种群蚁群算法在进化中容易出现早熟、停滞的现象,对算法进行了改进。在双种群蚁群分别独立进化、定期进行信息交换的基础上,提出一种新的蚁群优化算法,通过建立信息素扩散模型,并在每种蚁群的局部信息素更新上采用扩散模型,使蚂蚁更好的发挥了协作能力。以旅行商(Travel Salesman Problem,TSP)问题为例的仿真实验表明,该算法比基本双种群蚁群算法具有更好的收敛速度和寻优能力。The Basic Dual Population Ant Colony Optimization （BDPACO） often gets into precocity and stagnation during evolution. And an improved algorithm is presented to solve the problem. Based on dual population ant colony＇s independent evolution and regular information exchange,a new ant colony optimization is proposed. By setting up a pheromone diffusion model to update the local pheromone,ants improve the ability of cooperation. The simulation results on traveling salesman problem verified the algorithm has a higher convergence speed and better searching ability than the Basic Dual Population Ant Colony Optimization.

关 键 词：双种群 蚁群算法 信息素扩散 旅行商问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]