一种针对约束问题改进的人工蜂群算法  

An improved artificial bee algorithm for constrained problems

作　　者：李秀彦[1] 刘敬浩[1] 韩宏业[1] 

机构地区：[1]天津大学电子信息工程学院,天津300072

出　　处：《信息技术》2015年第5期125-128,共4页Information Technology

摘　　要：人工蜂群算法是基于自然界蜂群行为的一种算法,该算法已被广泛应用在不同的约束类问题,它能有效解决0-1背包等约束问题。文中提出了基于可行规则和多目标优化问题的改进人工蜂群算法。针对人工蜂群算法存在收敛速度低的缺点,提出一种针对约束问题改进的人工蜂群算法,通过改变一些参数提高算法的收敛速度。对著名的13个约束问题进行试验验证,实验表明改进的人工蜂群算法的最优解和平均解都优于原始的人工蜂群算法。Artificial bee colony （ABC） algorithm is an algorithm based on the behavior of bees in nature, the algorithm has been applied in different constraints problems, it can effectively solve 0 - 1 knapsack constraints problems and so on. This paper presents a practical rule-based optimization and multi-objective improved artificial bee colony algorithm. In order to solve the shortcomings of poor local search ability, low accuracy, easy to be trapped into local optimum that ABC algorithm exists, it presents an improved ABC by changing some parameters to improve the convergence speed, accuracy. Based on the famous 13 questions to test the constraint validation, the experiments show that the optimal solution and the average solution of improved ABC algorithm is superior to the original artificial bee colony algorithm.

关 键 词：人工蜂群算法 可行规则 多目标优化 约束问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]