基于粒子群的蚁群算法参数最优组合研究

A Research on the Optimal Combination of ACA Parameters Based on PSO

针对蚁群算法参数的不同取值对算法性能的影响,试图确定算法参数的最优组合,使算法性能最佳。在算法基本原理的基础上,分析各参数对算法性能的影响。提出确定蚁群算法参数最优组合的"两步走"策略,即先确定各参数的较优取值范围,再引入适应度函数并结合粒子群算法得到各参数的最优组合。仿真结果表明,提出的"两步走"策略能取得较好的效果,有利于蚁群算法的推广和应用。

Different value of parameters of ant colony algorithm （ACA） affects the performance of the algorithm. The paper tries to determine the optimal combination of algorithm parameters so as to gain the best algorithm performance. Based on the basic principle of the algorithm, effect of parameters on algorithm performance is analyzed. It also proposes a ＂two-step＂ strategy of the optimal combination which firstly determines a better range of parameter, then introduces reasonable function and gains the optimal combination of parameter with PSO algorithm. The simulating results show that this ＂two-step＂ strategy can achieve better effect and is helpful to the promotion and application of ACA.