一种增强局部搜索能力的改进人工蜂群算法

针对人工蜂群算法初始化群体分布不均匀和局部搜索能力弱的问题，本文提出了一种增强局部搜索能力的人工蜂群算法（ESABC）。首先，在种群初始化阶段采用高维洛伦兹混沌系统，得到遍历性好、有规律的初始群体，避免了随机初始化的盲目性。然后，采用基于对数函数的适应度评价方式，以增大种群个体间差异，减小选择压力，避免过早收敛。最后，在微分进化算法的启发下，提出了一种新的搜索策略，采用当前种群中的最佳个体来引导下一代的更新，以提高算法的局部搜索能力。通过对12个经典测试函数的仿真实验，并与其他经典的改进人工蜂群算法对比，结果表明:本文算法具有良好的寻优性能，无论在解的精度还是收敛速度方面效果都有所提高。

Improved artificial bee colony algorithm based on enhanced local search

The shortcomings of the artificial bee colony algorithm (ABC) are its uneven initial population distribution and weak local search. In this paper, we propose an ABC algorithm based on enhanced local search (ESABC). First, we employ a high-dimension chaotic system (Lorenz system) to obtain the ergodic and regular initial populations and to avoid the blindness of random initialization in the population initialization stage. Then, we introduce improved fitness evaluation methods based on the logarithmic function to increase the differences between individuals, reduce selection pressure, and avoid premature convergence. Lastly, inspired by the differential evolution algorithm, we propose a new search tactic that uses the best individual in the contemporary population to guide the renewal of the next generation, and thereby enhance the local search ability. We examined the performance of the proposed approach with 12 classic testing functions and compared the results with the basic and other ABCs. As documented in the experimental results, the proposed algorithm exhibits good optimization performance and can improve both the accuracy and convergence speed of the algorithm.