摘要
人工鱼群算法虽然具有不需要了解问题的特殊信息,能够寻找到一定的搜索方向,对初值和目标函数的要求不高等优点,但在人工鱼群算法的后期,有一部分人工鱼会聚集在局部最优周围或者处在漫无目的地随机游动状态,从而影响算法寻优的精度及收敛速度.针对这一不足,引入DNA交叉和DNA变异操作,提出一种基于DNA的改进人工鱼群算法.通过函数测试表明,该算法在搜索精度、可靠性、稳定性和鲁棒性4个性能上相比于基本人工鱼群算法更有效.
Artificial fish swarm algorithm has the advantage which has not to need special information of problems,is able to find a certain search direction,and has less demanding for initial and objective function. But in the late of artificial fish swarm algorithm,a part of the artificial fish will gather in the local optimal around or in the diffuse without destination random walk state,the algorithm accuracy and convergence rate of optimization will be af-fected. To solve this problem,DNA crossover and DNA mutation operation was introduced,and the improved arti-ficial fish swarm algorithm based on DNA calculation was proposed. Function test shows that the algorithm is more effective in the performances of search accuracy,reliability,stability and robustness compared with the basic artifi-cial fish swarm algorithm.
出处
《天津大学学报（自然科学与工程技术版）》
EI
CSCD
北大核心
2016年第6期581-588,共8页
Journal of Tianjin University：Science and Technology
基金
国家自然科学基金资助项目(61401307)
中国博士后科学基金资助项目(2014M561184)
天津市高等学校科技发展基金计划资助项目(20110709)
天津市应用基础与前沿技术研究计划资助项目(15JCYBJC17100)
中国物流学会资助项目(2014CSLKT3-16)
关键词
鱼群算法
DNA
交叉
变异
fish swarm algorithm
DNA
crossover
mutation
