带有征税算子的改进蚁群优化方法

摘要

Aiming at the disadvantage(premature convergence) of Ant Colony Algorithm(ACA),edified from the role of tax mechanisms of human society, the tax operator is presented to strengthen its global search ability. Tax operator restrains the rapid expansion of difference between pheromone in order to improve the solution. The preferences and the convergence of the tax operator is discussed in the paper. In the end,an example of Traveling Salesman Problem(TSP) is given in the paper,which is simulated by using basic ACA and improved ACA. The simulation results show that the tax operator has excellent global optimization properties,it can avoid premature convergence of ACO.%针对蚁群算法存在停滞现象的缺点,借鉴人类社会税收机制的作用,提出了能够强化其全局搜索能力的征税算子.征税算子通过抑止信息素差异急剧膨胀,以提高所得解的全局性.并对征税算子的参数设置以及收敛性问题进行讨论研究,最后将添加征税算子的蚁群算法与传统蚁群算法分别应用于旅行商问题(TSP)进行仿真实验.仿真结果表明,征税算子具有优良的全局优化性能,可抑制算法过早收敛于次优解,有效防止了停滞现象.