计算机应用 ›› 2005, Vol. 25 ›› Issue (07): 1641-1644.DOI: 10.3724/SP.J.1087.2005.01641

• 人工智能 • 上一篇    下一篇

用自适应的多种群蚁群算法求解频率分配问题

章春芳,陈崚,陈娟   

  1. 扬州大学 信息工程学院
  • 收稿日期:2005-01-04 修回日期:2005-02-27 出版日期:2005-07-01 发布日期:2005-07-01
  • 作者简介:章春芳(1982-),女,江苏南京人,硕士研究生,主要研究方向:算法设计、人工智能;陈崚(1951-),男,江苏宝应人,教授,博士生导师,主要研究方向:算法设计、人工智能;陈娟(1979-),女,江苏扬州人,硕士研究生,主要研究方向:算法设计、并行计算
  • 基金资助:

    国家自然科学基金资助项目(60074013);国家科技攻关计划项目(2003BA614A14);江苏省教育厅自然科学基金项目(02KJB520009)

Solving frequency assignment problem using an adaptive multiple-colony ant algorithm

ZHANG Chun-fang,CHEN Ling,CHEN Juan   

  1. School of Information Engineering, Yangzhou University
  • Received:2005-01-04 Revised:2005-02-27 Online:2005-07-01 Published:2005-07-01

摘要:

提出一种自适应的多种群蚁群算法求解移动通信中的频率分配问题。该算法改变了传统蚁群算法只有一个蚂蚁群体的做法,使用多个蚂蚁子群体同时进行优化处理。为每个蚂蚁子群体定义一个收敛系数,根据收敛系数来决定子群体内部的路径选择和信息量更新、子群体间的信息交流策略,同时采用自适应的信息更新策略以取得各蚂蚁子群体中解的多样性和收敛性之间的动态平衡。通过对固定频率分配和最小跨度频率分配问题进行仿真的实验,表明此算法不仅具有较强的全局收敛性,而且有更快的寻优速度。

关键词: 蚁群算法, 多种群, 收敛系数, 自适应, 频率分配

Abstract:

An adaptive multiple colony ant algorithm was presented to solve frenquency assignment problem of mobile communicaiton. Unlike the traditional ant colony algorithm which used only one ant colony, our algorithm used multiple ant colonies. For each ant colony, a coefficient of convergence was defined by which the ants adaptively could choose the path, update their local pheromone and exchange information between colonies. By using the adaptive strategy to update the pheromone, the balance between the diversity and convergence of every ant colony was kept. The simulation results on the fixed frequency assignment problem and minimal span frequency assignment problem show that our algorithm has global convergence and higher speed of optimization.

Key words: ant colony algorithm, multiple-colony, coefficient of convergence, adaption, frequency assignment

中图分类号: