中文摘要：

针对人工蜂群算法局部搜索能力弱及易陷入局部最优的缺点,提出了一种改进的人工蜂群算法。首先,雇佣蜂使用全局最优引导的搜索策略,且引导程度随个体试验次数（trial）自适应减小,以此平衡算法的全局和局部搜索能力;其次,观察蜂采用变异的异维学习策略,使算法的搜索具有跳跃性,以提高跳出局部最优的概率。对八个经典基准测试函数和CEC2013中八个复合基准函数的测试结果表明,与多种最近提出的类似算法相比,新算法在收敛速度和解的精度上均具有显著优势。

英文摘要：

In order to overcome standard artificial bee colony algorithm poor at exploitation and easily getting into local minima,this paper proposed an improved artificial bee colony algorithm. Firstly,it utilized a global optimum guided strategy whose level of guidance adaptively decreasing with the number of trials for employed bees to achieve a tradeoff between exploration and exploitation. Secondly,it used a mutated strategy with different dimensional learning for onlookers,so the search of algorithm with jumping coul improve the probability of escaping from the local minima. Through the experiment on 8 benchmark functions and 8 CEC2013 composition functions,the results show that the new algorithm performs significantly better than several recently proposed similar algorithms in terms of the convergence speed and the solution accuracy.