摘要
针对于标准的蚁群算法在问题规模较大情况下收敛速度慢、易早熟的问题,通过引入全局最优路径地图以及概率边计数值的方法,从而实现对于搜索路径当前最优值、全局最优值的启发性和历史的最优值的兼顾,提高了蚁群算法全局最优值查找能力,加快蚁群算法的收敛速度,提出了一个基于最优路径地图和概率边计数值的蚁群算法,通过理论证明改进的蚁群算法可以有效的提高蚁群算法对于旅行商问题的全局最优路径的查找能力以及对于最优值的收敛速度,通过对于不同旅行商问题实例的进行实验,验证了本文提出基于最优路径地图和概率边计数值的蚁群算法的合理性以及相应的理论证明的正确性.
Aiming to the problem of standard ant colony optimization (ACO) algorithm easy to prematurity and slow rate of conver- gence in large-scale situation, using the method of introduce of global best path map and probability edge count value which can im- prove iteration speed and the global optimal value searching capacity of the ant colony optimization algorithm , a improved ant colony optimization based on the probability edge count value and global best path map is proposed. Though theoretical derivation, the cor- rectness of improved ant colony optimization algorithm is proofed, and The correctness of theoretical derivation is verified by the ex- periment which is based on different TravelinR Salesman Problem ( TSP) instances.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第8期1831-1836,共6页
Journal of Chinese Computer Systems
基金
国家"八六三"自然科学基金项目(2013AA01A211)项目
关键词
蚁群算法
旅行商问题
概率边计数值
最优路径地图
ant colony optimization algorithm
travelling saleman problem
probability edge count value
global best path map