计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (2): 112-114.DOI: 10.3778/j.issn.1002-8331.2011.02.035

• 数据库、信号与信息处理 • 上一篇    下一篇

神经元网络模型的弱信号随机共振检测研究

耿丽硕,范影乐   

  1. 杭州电子科技大学 生物医学工程与仪器研究所,杭州 310018
  • 收稿日期:2009-08-25 修回日期:2009-10-26 出版日期:2011-01-11 发布日期:2011-01-11
  • 通讯作者: 耿丽硕

Research on neuron network of weak signal based on stochastic resonance detection

GENG Lishuo,FAN Yingle   

  1. Institute of Biomedical Engineering and Instrument,Hangzhou Dianzi University,Hangzhou 310018,China
  • Received:2009-08-25 Revised:2009-10-26 Online:2011-01-11 Published:2011-01-11
  • Contact: GENG Lishuo

摘要: 基于FitzHugh-Nagumo可兴奋细胞耦合后形成的神经元网络模型,对生物神经系统的弱周期信号随机共振检测机制进行研究。以加和网络的双层FHN神经元模型为例,对周期随机共振现象分别进行研究,并应用信噪比、互信息率对比评价方法,结合输出神经元动作电位的发放频率和幅值,从多个角度进行了定量和定性的描述和比较。实验结果表明,双层FHN神经元网络的随机共振响应优于单神经元的FHN模型,且具有更好的稳定性,可以在一定的噪声强度范围内对输入信号进行有效地检测。

关键词: FitzHugh-Nagumo, 神经元网络模型, 随机共振

Abstract: Based on neuron network model coupling by FitzHugh-Nagumo(FHN) exciting cells,the paper researches on stochastic resonance detection mechanism of weak periodic signal in biology nerve system.Through the example of double layers FHN neuron network model formed by summing structure,the stochastic resonance mechanism of periodic signal response are investigated.The adopted evaluate criterions are signal-to-noise ratio and mutual information rate,combined with the output electric spike velocity and amplitude of nerve cells.Qualitative and quantitative analysis from these points of view are done for the double layers FHN neuron network model.Results indicate that,the stochastic resonance response of the double layers FHN neuron network is better than the single FHN neuron model,and has better stability,and can be effectively detected for input signal at a wider range of noise intensity.

Key words: FitzHugh-Nagumo, neuron network model, stochastic resonance

中图分类号: