摘要： 为提高细菌觅食算法处理高维问题时的收敛速度及精度，提出一种基于粒子群优化算法和对立学习的细菌觅食算法PO-BFA。在种群初始化阶段采用对立学习取代随机初始化，在进化过程中利用对立学习进行种群动态跳跃，以提高算法的收敛速度，并以粒子移动代替细菌的趋化操作，由此省略细菌前进操作。基于6个高维Benchmark函数的实验结果表明，该算法的收敛速度和精度均优于同类算法。

关键词: 细菌觅食算法, 粒子群优化, 对立学习, 动态跳跃, 趋化

Abstract: To improve the convergence speed and accuracy of the basic Bacterial Foraging Algorithm(BFA) according to the high dimensional problems, this paper proposes a BFA based on Particle Swarm Optimization(PSO) and Opposition-based Learning(OBL), namely PO-BFA. It employs OBL for population initialization and for generation jumping, while uses the fly of particles same as PSO instead of bacteria chemotaxis, and omits the swim of the bacteria. Simulation results on six benchmark functions show that PO-BFA is superior to other kinds of BFA.

Key words: Bacterial Foraging Algorithm(BFA), Particle Swarm Optimization(PSO), Opposition-based Learning(OBL), dynamic jumping, chemotaxis

中图分类号: 

• TP301.6