基于信息素递减的蚁群算法

Ant Colony Algorithm Based on Pheromone Declining

通过对蚂蚁巡游路径的分析,发现经典蚁群算法在解决旅行商问题(TravelingSalesmanProblem,TSP)时的缺陷,在此基础上给出了新的信息素更新公式,提出了基于信息素递减的蚁群算法。新算法避免了蚂蚁在寻找最优解的过程中,由于禁忌表元素的逐渐增加而限制蚂蚁巡游路径选择的缺点,减少了巡游后期信息素对于后继蚂蚁的影响,提高了后继蚂蚁的巡游质量。通过具体的算例分析,表明此算法比传统的蚁群优化算法(AntColonyOptimization,ACO)算法有更快的收敛速度和非常好的稳定性。

Classical ant colony algorithm is found to be deficient for solving Traveling Salesman Problem （TSP） through analyzing ants＇ cruising route. Based on these, a new formula updating pheromone was introduced, and ant colony algorithm based on pheromone descending was proposed. The new algorithm avoids the defect that the gradually increased tabu table restricts the selection of ant cruising route during ants looking for the optimized solution, and it reduces the influence of pheromone on subsequent ants, enhances the subsequent ants＇ cruising quality. Experimental results on TSP show that the algorithm has faster convergence speed and great stability than that of classical Ant Colony Optimization （ACO） algorithm.