摘要： 提出了一种新的安全模型——模块化免疫神经网络模型，同时根椐模型中两大网络的作用机理，给出了它们之间的作用算法，最终形成了一个动态平衡的、自适应的、分布式的网络模型。实验表明模型有较高的安全数据识别率。

关键词: 拓扑结构；模型设计；免疫遗传算法

Abstract: In terms of the cluster particularity of neural network, this paper builds up a new security model——modular immune neural network, and constructs three algorithms from their mutual mechanisms, which forms a homeostic, auto-adaptable and distribution model. The experiment shows that this model has a better rate on data security classifying

Key words: Topology structure; Model design; Immune genetic algorithm