微弱正弦信号在有噪声的FN神经元模型中的传输特性研究

长期以来弄清神经系统中的信号是如何传输的一直是广大研究人员努力的目标.针对一种被普遍研究的神经元简化模型--FitzHugh-nagumo(FN)模型,采用二阶随机龙格-库塔算法分析了该模型对加性噪声和微弱正弦信号的响应特性.时域和频域的统计参数表明适当强度的噪声有利于信号的传输,存在随机共振现象,即与噪声强度关联的输出信噪比曲线为倒钟形;另外值得关注的是,与正弦信号频率关联的输出信噪比曲线也为倒钟形,分析可见正弦信号的无量纲频率在区间0.2～0.8时模型的输出信噪比最大,表明该神经元模型有频率敏感性,即更易于检测到该范围内的弱信号.上述结果与生物学的发现是一致的,将有助于进一步揭示周期信号在神经元中的传输方法,建立更加准确的神经元数学模型.

Transmission of Weak Sinusoidal Signal in the Noisy FN Neuron Model

To make out the signal processing of neuron and neural system,researchers have worked on it for scores of years.In this paper,the responses of FitzHugh-Nagumo(FN) neuron model,which is stimulated by additive Gaussian white noise and weak sinusoidal signal,is investigated via numerical simulations.The time and frequency domain results show that there is stochastic resonance and a dimensionless frequency sensitive range of 0.2-0.8.Thus the model can help to detect the sinusoidal signal.In addition,the frequen...