平衡搜索的改进人工蜂群算法

Improved Artificial Bee Colony Algorithm Based on Balanced Search

针对人工蜂群(ABC)算法局部搜索能力弱的问题,提出一种平衡搜索的人工蜂群算法(BSABC).首先,采用一种基于对数函数的的适应度评价方式,用于减小选择压力,在一定程度上避免陷入局部最优.其次,受微分进化算法的启发,提出一种新的搜索策略,通过当前最优个体指导进化方向,使候选解的产生倾向于当前最优解,同时避免陷入局部最优.对6个经典测试函数进行仿真实验,并与经典的改进人工蜂群算法对比测试,结果表明:所提出的算法在收敛速度和收敛精度上都有显著的提升.

Aim at the drawback of artificial bee colony(ABC)algorithm with weak local search capability,an artificial bee colony algorithm based on balanced search(BSABC)is proposed.Firstly,improved fitness evaluation methods based on the logarithmic function is introduced to minimize selection pressure and avoid to fall into local optimum to a certain extent.Secondly,enlightened by the differential evolution algorithm,a novel search strategy is proposed,which conducts the evolution direction of the candidate solution,tending to the current optimal solution,and at the same time avoiding to fall into the local optimum.The simulating experiments were conducted on a benchmark suite of 6 test functions,the results demonstrate that BSABC has significant enhancement in convergent speed and convergent accuracy compared with the basic ABC algorithm.