生物学杂志

• 技术方法 • 上一篇    下一篇

生物神经网络的建模与仿真

  

  1. 1.内蒙古工业大学 电力学院, 呼和浩特 010080; 2. 内蒙古机电控制重点实验室, 呼和浩特 010080
  • 出版日期:2016-06-18 发布日期:2016-06-18
  • 通讯作者: 董朝轶,教授,博士,研究方向为系统生物学、飞行器控制,E-mail: dongchaoyi@hotmail.com
  • 作者简介:刘剑钊,硕士,研究方向为生物信息处理,E-mail:liuxxxxliu@qq.com
  • 基金资助:
    国家自然科学基金地区基金项目(61364018);教育部留学回国人员科研启动基金(第45批);内蒙古自治区高等学校“青年科技英才计划”-青年科技领军人才;内蒙古工业大学青年学术骨干项目

Modeling and simulation of biological neural networks

  1. 1. College of Electric Power, Inner Mongolia University of Technology, Hohhot 010080;
    2. The Key Laboratory of Electromechanical Control, Inner Mongolia, Hohhot 010080, China
  • Online:2016-06-18 Published:2016-06-18

摘要:

生物神经网络系统是由许多的神经元之间通过突触相互连接起来,通过突触传递电信号,并且具有相当复杂的非线性网络系统。通过人工构造生物真实性的脉冲神经网络(Spiking Neural Networks, SNN)模型来模拟真实的神经元放电行为。首先,建立基于积分点火(Integrate-and-Fire, IF)机制的SNN模型;然后,确定模型中的参数,并对一个神经元和多个神经元网络进行仿真;最后,对比模型仿真的放电行为和真实神经元放电行为。仿真结果表明:基于IF模型的生物神经网络仿真能较好地逼近真实的生物神经网络。

关键词: 生物神经网络, 脉冲神经网络, 积分点火模型, 脉冲序列

Abstract:

Biological neural networks system is composed of a number of neurons through synapses between each other, transmitting electrical signals through the synapse, and has a very complex nonlinear network system. In this paper, the Spiking Neural Networks (SNN) model is constructed to simulate the real neuron discharge behavior. First, the SNN model is established based on the Integrate-and-Fire (IF) mechanism. Then, the parameters of the model are determined, and then the simulation of a neuron and a plurality of neurons is carried out. Finally, compare between model simulation and real neuron discharge behavior. The simulation result shows that the biological neural networks based on the IF model can be used to approximate the real biological neural networks.

Key words: biological neural networks, spiking neural networks, IF model, spiking series