一种自适应蚁群算法及其仿真研究

An Adaptive Ant Colony Optimization Algorithm and Simulation

蚁群算法是一种新型的进化算法,蚁群算法与其它进化算法同样存在易于限于局部最小点等缺陷。本文提出一种自适应的蚁群算法以克服上述缺陷。通过自适应地改变算法的挥发度等系数,本文中的算法可以在保证收敛速度的条件下提高解的全局性,通过对TSP问题的仿真证明本文中的算法相对与原始的蚁群算法收敛速度和解的性能都有一定的提高。

Ant Colony Optimization (ACO) algorithm is a new optimization algorithm, and has the limitation of stagnation. In this paper, the performance of ACO is improved by the adaptive variation of the parameter in the algorithm. The simulations for TSP problem show that the improved algorithm can find better paths at higher convergence speed.