摘要
蝙蝠算法(bat algorithm,BA)是受自然界中的蝙蝠通过回声定位进行搜寻、捕食行为的启发演变而来的一种新颖的群智能仿生优化算法。为了提高蝙蝠算法的收敛效率,把多种学习机制引入到蝙蝠优化算法中,通过将蝙蝠群体进行部落划分以及各部落间建立相互学习机制,使得内部局部搜索及全局最优信息能够在群体内传递。仿真结果表明,该算法切实提高了收敛效率。
Inspired by the echolocation behavior of bats, Bat algorithm is developed as a novel bionic swarm intelligence optimization method. In order to improve the global convergence, this paper combined multiple learning mechanism with BA and divided the tribe. It made the global optimal information were be transmitted within groups through the exchange of local search and global information. The simulation results Show that the proposed GDBA can indeed improve the global convergence.
出处
《计算机应用研究》
CSCD
北大核心
2015年第2期364-367,共4页
Application Research of Computers