【摘要】 针对人工蜂群算法在求解函数优化问题中存在收敛速度慢、易于陷入局部最优的缺陷,提出了一种基于当前最优解的混合变异算子的人工蜂群算法(artitificial bee colony algorithm based on hybrid mutation operator,HMABC).该算法中跟随蜂采用基于当前最优解的差分进化搜索策略,侦查蜂采用基于当前最优解的高斯变异侦查策略,通过变异增加种群的多样性,并且在当前最优解的引导下有效提高了算法收敛速度,避免其陷入局部最优.基于6种测试函数的仿真实验结果表明,提出的HMABC算法在收敛速度和求解精度方面均优于其他人工蜂群算法.更多还原

【Abstract】 In an artificial bee colony algorithm,it is prone to fall into local optimum and low convergence speed when solving function optimization problems. To alleviate this problem,a novel artificial bee colony algorithm based on hybrid mutation operator( HMABC) in the current optimal solution is proposed. With the help of current optimal solution,the differential evolution search strategy and Gaussian mutation detection strategy are adopted by followers and scouter,respectively. Under the guidance of current optimal solution,the diversity of population is increased via mutation,and the convergence speed is effectively improved with avoiding the local optimum. Simulation experimental results on six test functions show that,HMABC achieves satisfactory performance in the convergence speed and the accuracy when compared with some other artificial bee colony algorithms.更多还原