基于向量整体扰动的快速收敛人工蜂群算法

针对标准人工蜂群算法收敛速度慢和易陷入早熟收敛等问题,提出一种快速收敛人工蜂群算法。首先借助反向学习理论初始化种群来提高初始解的分布质量,并在雇佣蜂和跟随蜂阶段引入向量整体扰动搜索方程加快局部搜索;然后为了跳出局部最优,采用一种随机更新搜索策略来增加蜂群多样性以平衡全局探索和局部利用能力;最后通过八个标准测试函数的仿真实验,发现所提出的算法与几个改进的人工蜂群算法相比,具有更快的收敛速度且获得了更高的求解精度,验证了算法的优越性。

Fast convergent artificial bee colony algorithm based on full vector perturbation

To the defect of low searching speed and premature convergence frequently appeared in the standard artificial bee colony (ABC) algorithm, this paper proposed a fast convergent artificial bee colony algorithm based on full vector perturbation. Firstly, it employed opposition-based learning to initialize the population so as to improve the distributive quality of initial solutions. Secondly, it introduced search equation based on full vector perturbation to enhance local search during the first two phases of ABC. And then, in order to get out of the local minima and maintain the population diversity, it proposed random update search strategy for balancing between the ability of exploration and the one of exploitation. Lastly, experimental results tested on 8 benchmark functions show that the proposed algorithm reaches extremely faster convergence speed as well as better computational precision than other improved variants of ABC, which indicates that its performance is enhanced conspicuously.