计算机工程与应用 ›› 2019, Vol. 55 ›› Issue (3): 140-146.DOI: 10.3778/j.issn.1002-8331.1710-0255

• 模式识别与人工智能 • 上一篇    下一篇

差分进化引导趋化算子的烟花优化算法

刘  茜1,毛  力1,杨  弘2   

  1. 1.江南大学 物联网工程学院,江苏 无锡 214122
    2.中国水产科学院 淡水渔业研究中心,江苏 无锡 214081
  • 出版日期:2019-02-01 发布日期:2019-01-24

Fireworks Optimization Algorithm Based on Leading Differential Evolution Chemotaxis Operator

LIU Xi1, MAO Li1, YANG Hong2   

  1. 1.School of Internet of Things Engineering, Jiangnan University, Wuxi, Jiangsu 214122, China
    2.Freshwater Fisheries Research Center of Chinese Academy of Fishery Sciences, Wuxi, Jiangsu 214081, China
  • Online:2019-02-01 Published:2019-01-24

摘要: 标准烟花算法粒子间交流机制存在缺陷,且对最优点位置不在原点和原点附近时的目标函数求解能力差,对此提出差分进化引导趋化算子的烟花算法(BFA)。利用差分进化算法和趋化算子的局部搜索优势,在每一次迭代的过程中不断寻找这一代的最好个体,通过最优个体信息对局部粒子维度信息进行修改从而使得整个群体得到改善,8个标准和增加位置偏移的测试函数仿真结果表明,BFW相比于原始烟花算法(FA),粒子群算法和SPSO在寻优精度和寻优速度上有了较好的提高。

关键词: 烟花算法, 趋化因子, 差分进化, 函数优化, 全局寻优

Abstract: In order to solve the default of the fireworks algorithm inter-particle exchange mechanism and the disadvantage that the optimal position is not solved by the objective function near the origin and the origin come up with fireworks algorithm optimization with chemotaxis operator(BFW). Using the local search advantage of the chemotaxis operator to find the best individual in every iteration to improve the whole population’s search ability. The improved algorithm has been tested on 8 benchmark functions. The experimental results show that BFW has better behaviors in convergence accuracy and speed.

Key words: fireworks algorithm, chemotaxis operator, differential evolution, function optimization, global optimization