受启发的人工蜂群算法在全局优化问题中的应用  

Inspired Artificial Bee Colony Algorithm for Global Optimization Problems

作　　者：高卫峰[1] 刘三阳[1] 黄玲玲[1,2] 

机构地区：[1]西安电子科技大学理学院,陕西西安710071 [2]中国石油大学(华东)理学院,山东青岛266580

出　　处：《电子学报》2012年第12期2396-2403,共8页Acta Electronica Sinica

基　　金：国家自然科学基金(N.o60974082);中央高校基本科研业务费专项资金(No.K5051270002)

摘　　要：人工蜂群算法是最近提出的一种较有竞争力的优化技术.然而,它的搜索方程存在着探索能力强而开发能力弱的缺点.针对这一问题,受差分进化算法的启发,提出了一个改进的搜索方程.该搜索方程在最优解附近产生新的候选位置以便提高算法的开发能力.进一步,充分利用和平衡不同搜索方程的探索和开发能力,提出了一个改进的人工蜂群算法(简记为IABC).此外,为了提高算法的全局收敛速度,用反学习的初始化方法产生初始解.通过18个标准测试函数的仿真实验并与其他算法相比较,结果表明IABC算法具有良好的处理复杂数值优化问题的性能.Artificial bee colony （ABC） algorithm is a relatively novel optimization technique which has been shown to be competitive to other population-based algorithms. However, there is still an insufficiency in ABC regarding its solution search equa- tion, which is good at exploration but poor at exploitation. Inspired by differential evolution （DE）, we propose a modified solution search equation, which is based on that the bee searches only around the best solution of the previous iteration to improve the ex- ploitation. Furthermore, making full use of and balancing the exploration and the exploitation of different solution search equations, we present an improved ABC （IABC for short） algorithm. In addition, to enhance the global convergence, when producing the ini- tial population,the opposition-based learning method is employed.Experiments are conducted on a set of 18 benchmark functions. The results demonstrate good performance of the IABC algorithm in solving complex numerical optimization problems when com- pared with two ABC-based algorithms.

关 键 词：人工蜂群算法 差分进化算法 搜索方程 种群初始化 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]