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.
出处
《现代计算机》
2009年第12期65-67,96,共4页
Modern Computer
关键词
蚁群算法
信息素
路径优化
Ant Colony Optimization
Pheromone
Path Optimization
