基于多态自适应的蚁群算法研究

Research of Ant Colony Algorithm Based on Polymorphism and Adaptivity

针对基本蚁群算法易陷入局部最优,收敛速度慢等不足,提出了一种多态自适应蚁群算法:首先引入不同种类的蚁群,每种蚁群有各自不同的信息素调节机制;其次采用自适应调整信息素挥发因子的策略,并将各条寻优路径上可能的残留信息素数量限制在一个区间内,避免出现停滞现象。仿真结果验证了文章所提算法的可行性和有效性。

For basic ant colony algorithm （ACA） easy to fall in local best and slow converging speed, a polymorphic and adaptive ACA （PAACA） is proposed. First it imports different kinds of ants in which each kind has its own pheromone adjustment mechanism. Then it adopts the strategy of adaptively adjusting pheromone volatile factor, and the amount of possible residual pheromone on each searching optimal path is limited in an interval for avoiding stagnation phenomenon. The simulation results verify the feasibility and effectiveness of the proposed algorithm.