基于免疫算法的B-P网络权值设计  

Design of the Weights of B-P Networks Based on Immune Algorithm

作　　者：郝勇娜[1] 梁鸿生[1] 柴继河[1] 王凯[1] 

机构地区：[1]西安理工大学自动化与信息工程学院,陕西西安710048

出　　处：《昆明理工大学学报（理工版）》2004年第2期65-67,共3页Journal of Kunming University of Science and Technology(Natural Science Edition)

摘　　要：提出了一种基于免疫算法的B-P网络权值设计方法 ,用于实现B-P网络权值空间的搜索 .该免疫算法是借鉴生命科学中免疫的概念与理论提出的一种新的算法 ,算法的核心在于免疫算子的构造 ,即接种疫苗和免疫选择 .免疫算法具有较好的全局收敛性和很强的自适应环境的能力 .Based on the immune algorithm, an optimizing method is presented which is used to search the weights of B-P network. This immune algorithm--(IA) is proposed with analogy to the concept and the theory of immunity in biotic science. The core of the algorithm lies in constructing the immune operator that is realized by vaccination and immune selection. The immune algorithm has better global convergence and very strong self-adaptive ability to the environment. The experimental results prove the high efficiency of the immune algorithm in designing the weights of B-P network.

关 键 词：免疫算法 B-P网络 权值设计 适应度函数 免疫算子 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]