摘要
研究了蚁群算法在连续空间的函数寻优问题。通过修改蚂蚁信息素的留存方式和行走规则,定义了一个连续空间的蚁群算法。模拟蚂蚁用触角交流信息的过程提出了直接通信的学习机制,增强了蚂蚁的搜索能力。为了防止出现"早熟"现象,在局部搜索过程中嵌入了模拟退火的思想。同时为避免过大的残留信息,选择了新的信息增量计算函数。实例运算证明了算法的有效性。
An ant colony algorithm applied to continuous problems is proposed.This algorithm is defined by modifying both the ＂trail remaining＂ and the transfer rules.Based on the processes that ants exchange information through antennas,a novel study strategy ＂direct communication＂ is presented,which enhances the ants＇ ability to search the continuous space.In the meantime,a strategy of simulated annealing is embedded in the algorithm to improve the optimization performance and prevent ＂premature＂ phenomena during the local searching.In order to avoid the large residual information,the new information increment function is applied.Experimental results show that the proposed algorithm is effective.
出处
《计算机工程与应用》
CSCD
北大核心
2007年第23期74-76,共3页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of China under Grant No.60674092)
江苏省高技术研究项目(the High-Tech Research Program of Jiangsu Province of China under Grant NoBG2006010)
关键词
蚁群算法
连续空间寻优
学习机制
模拟退火
ant colony algorithm
continuous space optimization
study strategy
simulated annealing
