混沌免疫粒子群优化算法在BP网络训练中的应用  

BP network learning algorithm based on chaos immune particle swarm optimization

作　　者：张郭军[1] 韩琳[2] 徐坤[2] 

机构地区：[1]渭南师范学院计算机科学系,陕西渭南714000 [2]西安工程大学理学院,陕西西安710048

出　　处：《西安工程科技学院学报》2007年第4期484-488,共5页Journal of Xi an University of Engineering Science and Technology

基　　金：陕西省教育厅专项科研计划项目(05JK191);陕西省教育厅自然科学专项基金资助项目(06JK286)

摘　　要：将人工免疫系统中的克隆选择和混沌算法引入粒子群优化算法,提出一种混沌免疫粒子群优化算法.算法的主要特点是利用克隆和混沌变异等操作,提高收敛速度和种群的多样性.结合Iris分类问题,将新算法应用到BP网络的权值优化中,并和基于标准PSO算法的方法和单纯BP网络训练进行比较.实验结果表明,该算法性能优于所比较的两种算法,并且具有良好的收敛性和稳定性.In this paper, the mechanism of clonal selection of artificial immune system and chaos algorithm are involved into particle swarm optimization, and a chaos immune particle swarm optimization algorithm is proposed. The advantages are via clone and chaos mutation, the speed of convergent and diversity of the swarm are improved apparently. It was applied in BP neural network training combing with Iris-classify problem. The proposed algorithm compared with that of which was based on the standard PSO and BP algorithm. The results show that the proposed algorithm is superior to the other two algorithms with a better astringency and stability.

关 键 词：神经网络 粒子群优化算法 克隆选择 混沌算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]