摘要
蝙蝠算法(BA)是一种新的群智能优化算法。然而,BA算法的优化性能还不是十分完善,存在易陷入局部最优、早熟收敛等问题。针对BA算法的不足,提出一种具有记忆特征的改进蝙蝠算法,并考虑了由于时变或时滞引起的扰动问题。该算法中蝙蝠的前期搜索经验对后期搜索提供支持。实验结果表明,该算法具有较好的全局搜索能力和较快的收敛速度,能有效地克服早熟收敛问题。
Bat algorithm（ BA） is a new swarm intelligence optimisation algorithm.However,its optimisation performance still has some insufficiencies.BA algorithm has the phenomena of premature convergence and being easily fallen into local optimum.An improved bat algorithmwith memory characteristic（ MCBA） is proposed for improving these disadvantages,and the disturbance problems which are caused by time-variant or time-delay are also discussed.In this algorithm the search experience gained in earlier stage supports the searches in later stage.Experimental results show that the improved BA has better global search ability and a faster convergence speed,and can effectively overcome the problem of premature convergence.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第11期257-259,329,共4页
Computer Applications and Software
基金
国家自然科学基金项目(61074185)
广西自然科学基金项目(0832084)
广西高等学校科研项目(201202ZD032)
关键词
蝙蝠算法
记忆特征
扰动
Bat algorithm
Memory characteristic
Disturbance
