蚁群算法中蚂蚁更新方法之研究-计算机工程与应用.PDF

54 2011 ，47 （25 ） Computer Engineering and Applications 计算机工程与应用 蚁群算法中蚂蚁更新方法之研究 1 2 3 孟 非 ，李静宜 ，朱人杰 1 2 3 MENG Fei ，LI Jingyi ，ZHU Renjie 1.江苏科技大学 经济管理学院，江苏 镇江 212003 2.江苏科技大学 电子信息学院，江苏 镇江 212003 3.山东省邮电规划设计院有限公司，济南 250031 1.School of Economics and Management ，Jiangsu University of Science and Technology ，Zhenjiang ，Jiangsu 212003 ，China 2.School of Electronics and Information ，Jiangsu University of Science and Technology ，Zhenjiang ，Jiangsu 212003 ，China 3.Shandong Province Posts and Telecommunications Planning and Design Institute Limited ，Jinan 250031 ，China MENG Fei ，LI Jingyi ，ZHU Renjie.Study on updating algorithms in ant colony algorithm.Computer Engineering and Applications ，2011 ，47 （25 ）：54-57. Abstract ：Ant Colony Algorithm （ACA ）is a stochastic optimization algorithm inspired by the behavior of ant looking for food.However ，the standard ACA has some shortcomings ，such as premature convergence ，searching precision lowness and so forth.Based on the simulation of natural death process of 5% B-cell in biology clone selection ，this paper proposes 8 kinds of updating algorithms according to intergeneration differential ，theory of chaos ，principle of mutation respectively ，and selects the updated ants in terms of simulated annealing method.Numerical experiments show that the updating algorithm by using intergeneration differential and chaotic mutation is a good selection.Simultaneously ，the updating effect is perfect when the up- dated ants are about 20%.This algorithm can effectively overcome the premature problem and speed up the convergence. Key words ：ant colony algorithm ；clone selection ；chaos ；mutation ；simulated annealing 摘 要：蚁群算法是根据蚂蚁的觅食行为而提出的随机优化算法，但其存在早熟收敛和搜索精度低等问题。模拟生物克隆选择 中5%的B 细胞自然消亡过程，在蚁群算法进化过程中分别基于代间差分、混沌理论、变异原理等方法设计了8 种蚂蚁更新算法， 按照模拟退火方法进行更新后蚂蚁的选择。通过数值试验得出结论：基于代间差分和混沌变异的蚂蚁更新算法是一种很好的选 择，并且当性能较差的20%左右蚂蚁按