首页 | 本学科首页   官方微博 | 高级检索  
     

奖惩蚁群算法
引用本文:张飞君,高玮,汪磊.奖惩蚁群算法[J].计算机工程与应用,2010,46(10):44-47.
作者姓名:张飞君  高玮  汪磊
作者单位:1. 武汉工业学院研究生院,武汉,430023
2. 武汉工业学院土木系,武汉,430023
基金项目:国家自然科学基金No.40872187~~
摘    要:由于传统蚁群算法所采用的是随机概率搜索策略,收敛速度慢是其主要问题。为了提高算法的收敛速度,这里提出一种带奖惩策略的蚁群算法(PPACO)。新算法中,每次循环中发现的较优解都被挑选出来加以奖励,而普通解则被惩罚,这样就加快了较优路径和普通路径上信息素的差异;另外,为了不使这种差异对算法产生过多的影响,所有路径上的信息素都被限制在一定的范围τmin,τmax]内,同时,信息素的挥发系数被设为相对较高值。通过典型模拟实验证明,新算法对解决复杂组合优化问题非常有效。

关 键 词:蚁群算法  排序  奖励  惩罚
收稿时间:2008-10-6
修稿时间:2008-12-22  

Premium-penalty ant colony optimization
ZHANG Fei-jun,GAO Wei,WANG Lei.Premium-penalty ant colony optimization[J].Computer Engineering and Applications,2010,46(10):44-47.
Authors:ZHANG Fei-jun  GAO Wei  WANG Lei
Affiliation:ZHANG Fei-jun1,GAO Wei2,WANG Lei21.Department of Post-graduate,Wuhan Polytechnic University,Wuhan 430023,China 2.Department of Civil Engineering,China
Abstract:Due to the random probabilistic search strategy,the slow convergence is the main problem of the traditional Ant Colony Optimization(ACO).In order to improve the convergence of the algorithm,the Premium-Penalty Ant Colony Optimization(PPA-CO) is proposed.In this new algorithm,the better solutions found by the ants are awarded while the ordinary ones are punished. In order to counteract the polarization of pheromone values on all roads,the pheromone trails are limited to an interval τ_(min),τ_(max)] and the evaporation rate is set to a higher value.The results of the simulation experiments on several traveling salesman problems show that the PPACO is ranked among the best ACO for tackling the complicated combination optimization problems.
Keywords:ant colony optimization  rank  premium  penalty
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机工程与应用》浏览原始摘要信息
点击此处可从《计算机工程与应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号