一种具有动态自适应特征的改进的蚁群优化策略  

An Improved Ant Colony Optimization Strategy with Dynamic Self-Adaptive Characteristic

作　　者：朱百成[1] 周绍梅[1] 刘欣沂[2] 

机构地区：[1]南昌大学信息工程学院计算中心,南昌330031 [2]新余高等专科学校计算机系,新余338031

出　　处：《现代计算机》2009年第12期65-67,96,共4页Modern Computer

摘　　要：蚁群算法是一种有启发式搜索特性的仿生优化算法,在实际的应用中蚁群算法会表现出搜索速度慢、易于陷入局部最优以致算法停滞等缺点。提出一种改进的蚁群优化策略,当算法出现停滞时自适应地更改各路径上的局部信息素量大小,从而使它们的信息素差距动态地减小,以便于在后续的搜索中找出全局最优解。经过仿真实验表明,改进后的算法能发现更好的最优解。Ant colony optimization algorithm is a kind of bio-inspired optimization algorithm with a characteristic of heuristic search. But in practical application, the algorithm performances some shortcomings, such as slow search speed, easily to fall into local optimum algoritym for stagnation. Proposes an improved strategy of ant colony optimization, when the algorithm is stagnation, the pheromone is self-adaptive in every, route, and through this way to reduce the pheromone quantity difference among these routes. The simulations for TSP problem show that the improved algorithm can get better resuh.

关 键 词：蚁群算法 信息素 路径优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] Q959.837[自动化与计算机技术—计算机科学与技术]