一类反馈过程神经元网络模型及学习算法研究

Feedback procedure neural network and its algorithm

提出了一种带有反馈输入的过程式神经元网络模型,模型为三层结构,其隐层和输出层均为过程神经元。输入层完成连续信号的输入,隐层完成输入信号的空间聚合和向输出层逐点映射,并将输出信号逐点反馈到输入层;输出层完成隐层输出信号的时、空聚合运算和系统输出。在对权函数实施正交基展开的基础上给出了该模型的学习算法。仿真实验证明了该模型的有效性和可行性。

A novel neural network model named feedback procedure neural network (FPNN) is proposed. The FPNN has three layers, and its hidden layer and output layer are procedure neuron. Input layer accomplishes series signal input. Hidden layer accomplishes input signal space convergence and transfers input signal to output layer, then hidden layer transfers own output to input layer. Output layer accomplishes output signal convergence of space and time, and it accomplishes system output. A learning algorithm is presented based on weight function expanding by base function. The simulation experience proves availability of the model.