摘要
细菌觅食优化算法在搜索全局最优的过程中,可能出现细菌超出搜索空间的情况。针对这种现象,提出了7种受限制的边界条件和4种不受限制的边界条件,对出界细菌进行处理。通过单峰值函数和多峰值函数分别对这11种边界条件进行对比测试。结果表明,就收敛速度来看,不受限制的边界条件总体性能优于受限制的边界条件,而置于边缘的受限制的边界条件又稍微优于随机放置的情况。就搜索效果来看,对于单峰值函数,无形/衰减这种边界条件在寻优效果方面优于其他的边界条件,而且比较稳定,仅次于无形/衰减这种边界条件的是随机/吸收边界条件;对于多峰值函数,衰减边界条件更具优势。
In the process of searching global optimization,bacterial foraging optimization algorithm may be beyond the search space. In view of the phenomenon,seven restricted boundary conditions and four unlimited boundary conditions were proposed to deal with the outside bacterias. The eleven kinds of boundary conditions are separately compared by single peak functions and multimodal functions. The results indicate that on the convergence rate,the unlimited boundary conditions are superior to the restricted boundary conditions in the overall performance,and the conditions of placing particles on the edge of the restricted boundary are slightly better than those of placing particles randomly. On the search effectiveness,for single peak functions,the invisible / attenuation boundary condition is superior to the other boundary conditions in the effect of searching optimum,and the random / absorbing boundary condition is next to it. For multimodal functions,the attenuation boundary condition shows advantages.
出处
《计算机应用》
CSCD
北大核心
2015年第A02期111-113,154,共4页
journal of Computer Applications
关键词
细菌觅食优化算法
智能优化
边界条件
收敛
寻优效果
Bacterial Foraging Optimization（BFO） algorithm
intelligent optimization
boundary condition
convergence
effect of searching optimum
