遗传算法和蚁群算法优化TSP的设计与分析  

Design and Analysis of TSP Problem Based on Genetic Algorithm and Ant Colony Algorithm

作　　者：毕硕本[1] 董学士[1] 马燕[1] 

机构地区：[1]南京信息工程大学计算机与软件学院,南京210044

出　　处：《武汉理工大学学报》2010年第16期89-92,共4页Journal of Wuhan University of Technology

摘　　要：首先对TSP问题、遗传算法和蚁群算法做了简要介绍,然后给出了2种算法在求解TSP问题上基本的原理和步骤,并对该算法求解TSP问题进行设计分析与实验验证。结果表明：在实验条件下,种群数量定为5~15时,蚁群算法求解TSP问题的效果较好;种群数量定为城市数量的1~2.5倍时,遗传算法求解TSP问题的效果较优。This article firstly makes a brief introduction about TSP problem,genetic algorithm and ant colony algorithm,then gives the basic principles and steps of the two kinds of algorithms in solving the TSP problems,does designation analysis and experiments of the two kinds of algorithms for solving TSP problems,draws some useful conclusions： under the experimental conditions,while the population during 5 to 15,the ant colony algorithm for TSP problem is more effective;when the population is 1～2.5 times than cities,better results by using genetic algorithm for solving TSP.

关 键 词：遗传算法 蚁群算法 TSP 最短路径 组合优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]