摘要
本文提出加权值多态蚁群算法。在信息素初始化时加入权值,加大各条路径之间的信息素差异,利于蚂蚁快速进行路径选择;在概率选择过程中加入权值,提高蚂蚁搜索效率;采用了蚁周模型对信息素进行全局更新,并且设置了信息素最大值,避免算法陷入局部最优解。最后采用均匀分布的方法确定参数值,通过仿真实验结果表明,该方法在TSP问题中具有良好的稳定性和高效性。
This paper proposes weighted value polymorphic ant colony algorithm.Added weight q1 when pheromone initialization,increased pheromones differences between the paths,beneficial to the ants select path quickly.Added weight q2 when select probability, improve ants search effficiency.Adopted Ant-Cycle System,updated the pheromones and set up the max pheromones τmax,avoid the algorithm fall into local optima.Adopted evenly distribution method to determine parameter, simulation results show that the algorithm possesses good stability and efficiency.
出处
《软件工程》
2016年第4期1-4,共4页
Software Engineering
关键词
蚁群算法
权值
均匀分布
信息素
ants colony algorithm
weight
evenly distributed
pheromone
