求解函数优化问题的改进的人工蜂群算法  

Improved Artificial Bee Colony Algorithms for Function Optimization

作　　者：葛宇[1] 梁静[2] 王学平[3] 谢小川[1] 

机构地区：[1]四川师范大学基础教学学院,成都610068 [2]成都工业学院网络中心,成都610031 [3]四川师范大学数学与软件科学学院,成都610068

出　　处：《计算机科学》2013年第8期252-257,共6页Computer Science

基　　金：四川省教育厅项目(12ZB112)资助

摘　　要：为提高人工蜂群算法求解复杂函数优化问题的性能,分析了算法中侦察蜂逃逸行为的不足,并对其进行改进:定义了逃逸指标,使其能准确地反映个体状态对算法早熟的影响;重新设计选择机制,让侦察蜂不需要参数控制,能自适应地选择可能导致算法早熟收敛的个体执行逃逸操作;改进了逃逸算子,降低了逃逸操作的盲目性。通过9个典型测试问题的实验结果表明:在指定误差精度下,本改进算法均能有效收敛;同时与基本人工蜂群算法和已有的典型改进相比,本改进算法在收敛精度和速度上均有明显提高。说明提出的改进策略能有效提高算法求解复杂函数优化问题的能力。In order to enhance the performance of artificial bee colony algorithm in solving complex function optimiza- tion problems, this paper analysed the shortcoming of escape behavior of scout bees, and improved it. The improved al- gorithm defines escape index, making it precisely reflecting the effect of individual status on the premature convergence of algorithm, redesigns the selection scheme, making scout bees choosing individual escape operation that might result in algorithm premature convergence adaptively, improves the escape operator, reducing the blindness of escape operation. Nine typical experiments prove that the improved algorithm could converge efficiently under assignment convergence ac- curacy, and the improved algorithm could converge with more convergence accuracy and speed compared with basic arti- ficial colony algorithm and existing typical improved versions, thus proves the improved strategy proposed in this paper could boost capability of solving complex function optimization problems.

关 键 词：人工蜂群算法 早熟收敛 逃逸指标 选择机制 逃逸算子 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]