摘要
本文提出了一种多线程的高速收敛蚁群算法,该算法在MMAS基础上,采用多线程来实现其蚁群算法并行机制以减少寻路时间,同时结合粒子群算法中粒子位置转移的机制,采用一种新颖的最近邻居选择策略、并进行动态信息素更新策略,以保证在每次搜索中,都能迅速向较优解靠拢.同时,还采取了一种局部变异策略,以对每次搜索的结果进行优化.
This paper present a quickly convergent version of the ACO algorithm.On the basis of the max-min ant system,the algorithm uses the means of multi-thread to realize the mechanism of parallel and to shorten the time of finding the route.At the same time,a mechanism in transformation of location of particle swarm optimization and closet neighbor strategy along with the dynamic pheromone updating are adopted to ensure that each ant could quickly access to the nice route.Meanwhile,a local mutation mechanism is employed to optimize the search results of each circulation.
出处
《电脑知识与技术（过刊）》
2007年第14期531-533,共3页
Computer Knowledge and Technology
基金
福建省青年科技人才创新基金(2005J011),福州大学科技发展基金(2005-XQ-22).
关键词
多线程
粒子群算法
最近邻居
动态信息素更新
变异策略
multi-thread
particle swarm optimization
nearest neighbor
dynamic pheromone updating
mutation algorithm
