趋药性人工蜂群算法训练神经网络研究  

Artificial Bee Colony with Chemotaxis Behavior for Training Artificial Neural Network

作　　者：林晓宇[1] 钟一文[1] 王爱荣[1] 

机构地区：[1]福建农林大学计算机与信息学院,福建福州350002

出　　处：《广西师范大学学报（自然科学版）》2011年第3期120-124,共5页Journal of Guangxi Normal University:Natural Science Edition

基　　金：国家自然科学基金资助项目(30800713);福建省自然科学基金资助项目(2008J0316);福建省教育厅科技项目(JA10118)

摘　　要：本文设计了一种复合人工蜂群算法,将细菌觅食优化算法中的趋药性行为引入到人工蜂群算法中,使得引领蜂和观察蜂进行局部探查时像细菌觅食时那样翻转到有利的方向上进行游动;另一方面,让蜜蜂尝试在所有维度产生扰动并择优选择,这两种策略大大增强了人工蜂群算法的局部探查能力。将此算法应用于训练人工神经网络,实验表明改进后的算法性能比人工蜂群算法和粒子群算法有很大提高。A hybird artificial bee colony named ABC-CB（artificial bee colony with chemotaxis behavior） is presented to imporve the local exploitation ability of ABC.Chemotaxis behavior of bacterial foraging optimization algorithm is embedded into ABC in exploitation process to make employed bees and onlookers tumble to the best dimension and swim for several steps just like what the bacteria do while foraging.On the other hand,bees try all dimensions and select the best one to move.The simulation experiments carried out on the training of artificial neural network show that the proposed algorithm has better performance than bee colony optimization algorithm and particle swarm optimization algorithm.

关 键 词：趋药性 人工蜂群 人工神经网络 细菌觅食优化 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]