论文部分内容阅读
提出了振荡型混沌神经网络的结构模型和动力方程 ,证明了该网络的稳定性和 Hopf分支的必要条件 ,得到了在三维模式空间中的轨迹 ,利用功率谱密度分析了重迭函数的波形 .试验结果表明输入模式在靠近和远离记忆模式两种情况时 ,网络具有不同的信息处理能力 .该网络在时空模式的动力信息处理和智能信息处理系统中有很重要的应用价值
The structural model and dynamic equation of oscillating chaotic neural network are presented. The stability of the network and the necessary conditions of Hopf branching are proved. The trajectory in three-dimensional model space is obtained. The waveform of overlap function is analyzed by power spectral density. The experimental results show that the network has different information processing capabilities when the input mode is close to and far from the memory mode.The network has very important application value in the dynamic information processing and intelligent information processing system of space-time mode