蚁群算法(ACS)是一种新型的分布式模拟进化算法,它有较强的解搜索能力、很好的适应性和鲁棒性等,但如果算法中各参数选择不当,则会使算法的运行时间变长,或者陷于局部最优,达到停滞状态。恰当的参数选择,可以使蚁群算法有较好的性能.较快地收敛到全局较优解。以TSP问题为例,通过采用不同参数匹配进行优化的数值实验,分析了算法中参数α,β,ρ对算法性能的影响,给出了一定指导性的建议。
Ant Colony System(ACS) algorithm has,if the appropriate parameters are selected,a rather optimistic performance,and can reach the global optimistic solution effficiently.There are many positive characteristics in ACS algorithm.h can,for instance, adapt to the new circumstances swiftly,and it is a robust algorithm.However,if the parameters are wrongly chosen,all these merits are gone.In this paper,lots of experiments are run to test how the parameters α,β,ρ work,and some useful suggestions are given.