增强寻优能力的改进人工蜂群算法  

Modified Artificial Bee Colony Algorithm for Search Ability Enhancement

作　　者：易正俊[1] 韩晓晶[1] 

机构地区：[1]重庆大学数学与统计学院,重庆401331

出　　处：《数据采集与处理》2013年第6期761-769,共9页Journal of Data Acquisition and Processing

基　　金：国家自然科学基金(69674012)资助项目;重庆市科技攻关计划(CSTC2009AC3037)资助项目

摘　　要：针对人工蜂群算法易陷入局部最优问题,首先将蜂群算法中的跟随蜂数量翻倍,用其一半采用轮盘赌选择机制更新,保持蜂群沿着蜜源浓度大的方向进化;增加的一半跟随蜂采用反向的轮盘赌选择机制,用以维持种群的多样性。其次,将所有未更新计数器次数大于阈值的蜜源对应的引领蜂变成侦察蜂,并对相应蜜源进行更新搜索。最后,求出每一轮迭代后所得蜜源的中心位置,通过中心位置与每个蜜源所在的邻域内产生一新解,再比较适应度值的大小,选择优者。仿真实验证明,该改进算法具有更高的收敛精度和很好的鲁棒性,且增加了算法跳出局部最优的机会,增强了蜂群算法的寻优能力,具有更好的优化性能。Considering the problem that the artificial bee colony is inclined to fall into local opti- mum, firstly the number of onlooker bees is firstly doubled, the roulette wheel selection mech- anism is adopted to update by half bees for keeping evolution along the high concentration of nectar sources; and the roulette wheel selection mechanism of the reverse is used by the other half to maintain population diversity. Secondly, employed bees is transformed corresponding to all the nectar sources which the times of counter without updating is greater than the threshold value into scout bees, then these nectar sources are updated. Thirdly, the center position of nectar sources is calculated, which are obtained after each iteration. A new solution is produced in the neighborhood, formed by the center position and every nectar source. Then the values of fitness are compared to choose the optimum. Finally, experiments show that the improved col- ony algorithm has better convergence precision and good robustness. It increases the chance of the algorithm to jump out of local optimum, and enhances the ability of search, thus showing a better optimization performance.

关 键 词：人工蜂群算法 轮盘赌 适应度值 中心位置 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]