摘要
蚁群优化是一种元启发式的随机搜索技术.信息素是蚁群进行交流并实现群集智能的媒介,所以信息素的更新策略一直是蚁群算法中的一个研究热点.针对信息素扩散的耦合特征,提出一种基于信息素扩散模型解耦控制策略的蚁群算法.对信息素扩散模型进行改善,建立以蚂蚁经过的路径(直线段)为信源的信息素扩散模型,通过分析信息素扩散浓度场的耦合性,引入去耦控制策略来修正信息素的更新公式,大量TSP(traveling salesman prob-lem)问题的实验表明:该算法不仅能获得更好的解,而且能加快算法的收敛速度.
Ant colony optimization （ACO） is a meta-heuristic search technique. Pheromones are an important media ants use to communicate with each other and implement swarm intelligence. Thus research on pheromone updating strategies is a hotspot in ACO. A new decoupling control strategy model of pheromone diffusion is proposed based on the coupling characteristic of pheromone diffusion. First, the algorithm sets up a new pheromone diffusion model with the path that the ant travels as the pheromone source. Then, according to the coupling degree of the concentration field of pheromone diffusion, a decoupling control strategy is employed to revise the pheromone updating formula. Experimental results for many TSP problems demonstrate that the proposed algorithm can not only generate better solutions but also accelerate the speed of convergence.
出处
《智能系统学报》
2007年第4期1-8,共8页
CAAI Transactions on Intelligent Systems
基金
国家自然科学基金资助项目(60496322)
北京市教育委员会科技发展资助项目(KM200610005020)
北京市委组织部优秀人才培养资助项目(20061C0501500190)
关键词
蚁群算法
扩散模型
耦合性
解耦控制策略
ant colony optimization
diffusion model
coupling characteristic
decoupling control strategy
