基于最近邻居选择、信息素动态更新和局部启发搜索的蚁群算法

ACO Algorithm Based on the Nearest Neighbor Node Choosing, Dynamic Pheromone Updating and Local Searching

文章使用最近邻居选择、信息素动态更新和局部启发搜索法对MMAS算法进行优化,得出NDLACO算法.此算法运用于解CVRP问题时,取得了较好的效果.在关于参数值的问题上取得了一定的成效,也有效地解决了蚁群算法的收敛过快和早熟、停滞问题.

Although Ant Colony Optimization （ACO） algorithm has solved many NP-hard problems successfully, some parameters （α,β,ρ） must be changed for the sake of getting optimal results when different problems are solved. This paper optimizes MMAS by using the strategies of the nearest neighbor node choosing, dynamic pheromone updating and local searching and obtains NDLACO algorithm, which shows some advantages in dealing with some instances of CVRP, parameters, and ACO＇s shortcomings such as its fast convergence and stagnation.