基于自适应多态免疫蚁群算法的TSP求解  

Solving TSP based on adaptive polymorphic immune ant colony algorithm

作　　者：吴建辉[1,2] 章兢[1] 刘朝华[2] 

机构地区：[1]湖南大学电气与信息工程学院,长沙410082 [2]湖南大学计算机与通信学院,长沙410082

出　　处：《计算机应用研究》2010年第5期1653-1658,共6页Application Research of Computers

基　　金：国家自然科学基金资助项目(60634020;60874096);湖南省自然科学基金资助项目(07JJ3126)

摘　　要：针对标准蚁群算法易于出现早熟停滞现象,提出了一种自适应多态免疫蚁群算法(adaptive polymorphic immune antcolony algorithm,PIACA)。通过设置多种状态蚁群及引入自适应多态蚁群竞争机制,PIACA算法能有效抑制收敛过程中的早熟停滞现象。将禁忌表中每只蚂蚁走过的路径视为抗体,对抗体运用局部最优搜索算法和免疫克隆选择算法进行高效优化,提高了解的质量。针对TSP实验结果表明,该算法在收敛速度及求解精度上均取得到了较好的效果。Aiming at the phenomenon that standard ant colony algorithm was easy to occur premature and stagnation,this paper proposed an adaptive polymorphic immune ant colony algorithm（PIACA）.PIACA algorithm could suppress premature and stagnation behavior of the convergence process through setting multiple ant groups with different state and the introduction of adaptive polymorphic ant colony competition mechanism. Paths which ants traverse were regarded as antibodies, antibodies in the tabu table were effectively optimized by local optimization searching algorithm and immune clone selection algorithm, the quality of solution was significantly improved through above operations.Simulation test for traveling salesman problem（TSP） illustrates that PIACA algorithm has a remarkable quality of convergent precision and the convergent velocity.

关 键 词：自适应 多态 蚁群算法 免疫克隆选择 旅行商问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]