【摘要】 利用ＢＰ神经网络，对非线性系统产生的混沌序列进行学习，逼近该非线性系统的映射特征，从而使自身具有混沌特性，成为具备混沌输出能力的神经网络，建立了基于神经网络的混沌序列产生模型（ＣＧＮＮ）。此方法利用具有学习能力的神经网络，它不需要针对某一种非线性映射设计单一的系统结构．并利用ＤＳＰ技术在ＣＧＮＮ基础上，制成了混沌神经网络协处理机插板，将其产生的混沌序列作为密钥，用于信息的保密通信。更多还原

【Abstract】 Back propagation neural network and chaos dynamics are combined in this paper. The global-optimum and dynamic-adjusting neural network is trained by the chaos series from non-linear system so as to approach the mapping feature of that system. So, a chaos generation scheme based on the neural network is proposed. By using nonlinear neural network with learning ability, this scheme doesn’t have to be designed to be a single system structure for a certain nonlinear chaotic map. At last, using DSP technology, a Chaos Neural Network Co-processer Flashboard is made based on the CGNN. The chaos is used as a encrypeted key in the encryption telecommunication.更多还原