全局最优引导的差分演化二进制人工蜂群算法  

Differential evolution binary artificial bee colony algorithm based on global best

作　　者：刘婷[1,2] 张立毅[1,2] 鲍韦韦[3] 邹康[3] 

机构地区：[1]天津商业大学信息工程学院,天津300134 [2]天津大学电子信息工程学院,天津300072 [3]天津工业大学电子与信息工程学院,天津300387

出　　处：《计算机工程与应用》2013年第6期43-47,共5页Computer Engineering and Applications

摘　　要：针对基本二进制人工蜂群算法开采能力弱、收敛速度慢的缺点,提出一种全局最优引导的差分二进制人工蜂群算法。算法仿照粒子群优化,将全局最优参数引入二进制人工蜂群算法中以提高开采能力;同时受差分演化算法中"交叉"操作的启发,提出多维邻域搜索方式,加快收敛速度。采用0-1背包问题进行仿真,实验结果表明与传统算法相比,提出算法不仅寻优能力增强且收敛速度明显提高。对于10维背包问题,提出算法的收敛速度比基本二进制人工蜂群算法提高近10倍。The Basic Binary Artificial Bee Colony（BABC） algorithm has the disadvantages of poor exploitation and slow conver- gence speed. According to the defects, a differential evolution binary artificial bee colony algorithm based on global best is proposed. Referring to particle swarm optimization, global best parameter is incorporated into BABC algorithm to raise the exploitation capacity. Inspired by crossover operation in differential evolution algorithm, multidimensional neighborhood search strategy is applied to improve convergence speed. The 0-1 knapsack problem is simulated. The simulation results show that compared with the traditional algorithm, the proposed algorithm＇ s search ability is enhanced and its convergence speed is improved obviously. For 1 O-dimension knapsack problem, the convergence speed of the proposed algorithm is faster than that of basic BABC algorithm nearly 10 times.

关 键 词：基本二进制人工蜂群算法 粒子群优化 差分演化 全局最优 多维邻域搜索 0-1背包 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]