遗传蚁群禁忌融合算法的研究  

Research on genetic ant colony taboo fusion algorithm

作　　者：李亚鹏[1,2] 丁文霞[1] 

机构地区：[1]国防科学技术大学电子科学与工程学院,湖南长沙410073 [2]武警医学院临床医学系,天津300162

出　　处：《现代电子技术》2012年第7期77-80,83,共5页Modern Electronics Technique

摘　　要：通过对遗传算法、蚁群算法和禁忌搜索算法三种算法的分析研究,针对其各自优缺点,提出一种融合遗传算法、蚁群算法和禁忌搜索算法的融合算法。融合算法是采用遗传算法生成初始信息素分布,利用蚁群算法快速求精确解,同时将遗传禁忌算子引入到蚁群算法的每轮迭代中,有效解决了蚁群系统初始信息素匮乏、易陷入局部最优和收敛速度慢的缺点,实现优势互补。通过NP-hard30问题仿真实验,结果显示算法具有良好的寻优能力和寻优效率。A new fusion of genetic algorithm,ant colony algorithm and taboo search hybrid algorithm is presented.Fusion algorithm used genetic algorithm to generate initial pheromone distribution,used ant colony algorithm for fast finding exact solutions,while genetic taboo operator was introduced into the ant colony algorithm in each iteration,the ant colony system initialization pheromone shortage,easily falling into local optimum and slow convergence were effectively solved,the complementary advantages were implemented.Through NP-hard30 simulation,results show that the algorithm has good optimization ability and searching efficiency.

关 键 词：遗传算法 蚁群算法 禁忌搜索算法 融合算法 仿真实验 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]