项目说明报告
基于蚁群算法路由选择可视化动态模拟
Visul Simulation of Routing Selsect based on Ant Colony Algorithms
路由选择是一种基于网络层的协议，而所有流行的网络层路由选择协议都是
基于以下两种典型的分布式算法之一：距离向量路由算法和链路状态路由算法。
组合优化问题是人们在工程技术、科学研究和经济管理等众多领域经常遇到的问
题，其中许多问题如旅行商问题、0-1背包问题、图着色问题、装箱问题等，都
被证明为NP-困难问题。用确定性的优化算法求NP完全问题的最优解，其计算时
间使人难以忍受或因问题的高难度而使其计算时间随问题规模的增加以指数速
度延长。用近似算法如启发式算法求解得到的近似解不能保证其可行性和最优性，
甚至无法知道所得解同最优解的近似程度。因而在求解大规模组合优化问题时，
传统的优化算法就显得无能为力了。在过去的10多年，蚁群算法（ACO）的研究
和应用取得了很大的进展，大量结果证明了算法的有效性和在某些领域的优势。
蚁群算法是一种新型的模拟进化算法, 研究表明该算法具有并行性, 鲁棒性等优
良性质。本文阐述了蚁群算法的原理,详细的说明了蚂蚁算法中各个功能模块，
并介绍了该算法在理论和实际问题中的应用, 并对其前景进行了展望。
蚁群算法 信息素 仿真
Abstract
Whether it is one based on Internet agreement for route not to choose, and all
Internet route that prevail choose agreement on the basis of the following two typical
distributed algorithm one of. Is it optimize problem people in engineering , scientific
research , economic management numerous problem that field run into often to make
up, among them a lot of question if knapsack issue , issue of businessman in the travel
industry and of TSP , pursue painted question , case issue ,etc., proved as 6WF
difficult problem. Ask the solving optimumly of JSP complete problem with the
deterministic optimization algorithm, calculation its time make people to be
insufferable making their calculation time up to increase , issue of scale lengthen so as
to index speed because the question is highly difficult. If heuristic algorithm is it solve
receive approximate solution can the assurance feasibility and getting optimum their
to ask with algorithm of similar toing, it is even unable to know incomes and solve
and solve optimumly to be similar to the degree. Therefore while asking and solving
and making the question of optimizing up on a large scale, the traditional optimization
algorithm seems powerless . From vectorial route algorithm, algorithm of route and
state of chain The researches and applications on ACO algorithm have made great
progresses in the past more than ten years. A number of results prove the validity of
the algorithm and its advantages in some fields. ACO algorithm whether one
new-type simulation evolve the algorithm , studies have shown this algorithm has
walking abreast nature, fine nature such as being stupid and excellent. This text has
explain ant's principle of one group of algorithms, has introduced this application in
the theory and practical problem of algorithm, and has looked forward to its
prospect .
Keyword: Ant Colony Optimization algorithm Pheromone Simulation
- i -
目 录
前言...............................................................1
第 1 章 绪论 ......................................................2
1.1 路由选择的意义 ............................................2
1.1.1 路由选择技术的组成....................................2
1.1.2 路由算法设计目标......................................3
1.1.3 路由算法的分类........................................4
1.1.4 路由算法衡量的标准...................................4
1.2.目前常用的路由算法 .........................................5
1.2.1 最短路径算法..........................................5
第 2 章 蚁群算法的基本原理 ........................................7
2.1 蚂蚁算法的产生..............................................7
2.2 蚂蚁算法的算法思想 .........................................7
2.3 蚁群算法原理................................................8
2.4 蚁群算法的应用 ............................................12
2.4.1 蚂蚁算法在电信网动态路由优化中的应用 .................12
2.4.2 蚂蚁算法在组合优化中的应用 ...........................12
2.5 蚂蚁算法的未来发展 .......................................12
2.5.1 MMAS ( Max2Min ant system) 最大最小蚁群算法..........12
2.5.2 具有变异特征的蚁群算法...............................12
2.5.3 自适应蚁群算法.......................................13
2.5.4 大规模集成电路综合布线 ...............................13
2.5.5 电信网络路由 .........................................13
第 3 章 开发工具 ..................................................14
3.1 软件环境...................................................14
3.2 其他资料...................................................14
3.3 Java 的简单介绍 ..........................................14
3.3.1 网络时代的需要.......................................14
3.3.2 Internet 的普及 ......................................14
3.3.3 跨平台可移植性的要求.................................14
3.4 Java 的主要特点 ...........................................15
3.4.1 简单性...............................................15
3.4.2 安全性...............................................15
3.4.3 面向对象性...........................................15
3.4.4 可靠性...............................................16
第 4 章 具体的功能结构 ............................................17
4.1 系统的结构总框图 ..........................................17
4.2 蚂蚁算法的主要步骤 ........................................18
第 5 章 系统的实现 ...............................................25
5.1 蚁群算法的实现结果.........................................25
第 6 章 算法的不足和改进 .......................................29
6.1 算法的不足 ................................................29
- ii -
6.2 算法的改进 ................................................30
6.2.1 信息素更新参数微调 ...................................30
6.2.2 全局调整.............................................31
6.2.3 信息素值微调.........................................31
6.3 一种先进的蚂蚁算法——智能蚂蚁算法.........................31
6.3.1 取消外激素...........................................31
6.3.2 自动调节选择最优路径的比例...........................32
5.6.3 选择目标城市的依据..................................32
6.3.4 引入扰动 .............................................32
6.4 蚂蚁算法的展望 ............................................33
结束语............................................................34
参考文献..........................................................35
