一个脉冲发放型人工神经元

An Artificial Neural Net Firing Impulsse

本文首先总结了生物神经元的一些基本特性,对调节联接的霍伯(Hebb)规则提出了新的见解,得出生物神经元的总和过程是间歇式的结论,通过对支配神经元活动的Hodgkin-Huxley方程进行合理简化得到一个具有类似于生物神经元动作方式的人工神经元;给出了这一模型的电路实现和解析解 ;最后,研究了该模型组成的多层网络实现异或分类问题.该模型具有生物神经元的主要特性,如总和效应、潜伏期和不应期等,并且兼顾到在工程上其容易由硬件、软件实现等问题.

The paper first summarizes some basic performances of the neuron, gives the new explanation for the Hebb 's law modifying the strength of the synapse control, obtains the conclusion that the timespace summation of neuron is only intermittant working. An artificial neural net is derived by directly simplifying the Hodgkin-Huxley equations governing the neuro. Lastly, an example is given that the multi-layer neural networks can realize the exclusive or assortment. This model has the primary performances of neuron, for example, the summafion, the latency and the refractory period.In addition, the model can be easily implemented by hardware and software in engineering.