基于MapReduce的混合蚁群算法研究  

A Hybrid Ant Colony Algorithm Based on MapReduce

作　　者：蔡明[1] 左勇安 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《计算机与现代化》2016年第10期6-9,共4页Computer and Modernization

摘　　要：传统的蚁群算法在收敛速度上较慢且容易导致局部最优解,本文提出一种基于双模式的混合蚁群算法,即在算法的每次迭代中有比例地选择其中一种模式来获得蚂蚁的最优路径,可以实现在相对较少的时间内寻找出最优路径,且避免陷入局部最优解。由于蚁群算法天然具有并行化的特性,本文将混合蚁群算法与MapReduce结合,大大缩短了算法的执行时间。实验结果表明,基于MapReduce的混合蚁群算法可以实现在相对较少的时间内寻找出较优的路径。The traditional ant colony algorithm has a slow rate of convergence and is easy to result in local optimal solution. This paper raises a new hybrid ant colony algorithm, which is based on a mixed mode of elite mode and normal mode. The algorithm selects a mode proportionally in each iteration to obtain the optimal path. In this way, we are able to find the optimal path in a less time and avoid falling into local optimal solution. Because of the parallel property of ant colony algorithm, it＇ s feasible to use MapReduce to run it. Experimental results show that the MapReduce-based hybrid ant colony algorithm can find out the optimum path in a relatively less time.

关 键 词：蚁群算法 混合蚁群算法 MAPREDUCE 云计算 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]