改进的人工鱼群算法  

Improved Artificial Fish-school Algorithm

作　　者：范玉军[1] 王冬冬[2] 孙明明[3] 

机构地区：[1]德州职业技术学院基础部,山东德州253000 [2]青岛滨海学院本科部,青岛266555 [3]广西民族大学数学与计算机科学学院,南宁530006

出　　处：《重庆师范大学学报（自然科学版）》2007年第3期23-26,共4页Journal of Chongqing Normal University:Natural Science

基　　金：广西自然科学基金项目(No.0542048);广西民族大学重大科研基金项目(No.0609013)

摘　　要：通过对人工鱼群算法的研究,给出了改进的人工鱼群算法。采用最优个体保留策略对觅食行为进行改进,防止群体中最优个体的退化;给出加速个体局部搜索方法,改进算法中的聚群行为和追尾行为,使全局最优值更快地突现出来;根据双射的定义和性质,在不影响最终寻优结果的情况下对问题的搜索域进行"缩小",从而加速了全局搜索。仿真结果表明改进的人工鱼群算法具有求解精度高、寻优成功率高、收敛速度快、算法稳定等优点。In this paper the Improved Artificial Fish-school Algorithm is proposed based on the study of the Artificial Fish-school Algorithm. The preying behavior is improved by introducing the strategy of keeping the best individual, and this method prevents the degenerating of the best individual in colony. The method of accelerating individual local searching is put forward, and it is used to improve the swarming behavior and fish＇s following behavior in order to make the global optimal value to be shown faster. Without influence on the final results, the searching area is ＂reduced＂ based on the definition and properties of bijection, so that global searching is accelerated. Several computer simulation results show that the Improved Artificial Fish-school Algorithm has some advantages such as higher precision of solution, higher efficiency of optimization, faster convergence rate, and better stabilization etc.

关 键 词：人工鱼群算法 搜索域 双射 改进的人工鱼群算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]