en
×

分享给微信好友或者朋友圈

使用微信“扫一扫”功能。
通讯作者:

韩芳,E-mail:yadiahan@dhu.edu.cn

中图分类号:Q42

文献标识码:A

文章编号:1672-6553-2023-21(9)-067-007

DOI:10.6052/1672-6553-2022-033

参考文献 1
KURAMOTO Y,BATTOGTOKH D.Coexistence of coherence and incoherence in nonlocally coupled phase oscillators [J].Nonlinear Phenomena in Complex Systems,2002,5(4):380-385.
参考文献 2
ABRAMS D M,STROGATZ S H.Chimera states for coupled oscillators [J].Physical Review Letters,2004,93(17):174102.
参考文献 3
OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.
参考文献 4
PAVLOVA O N,PAVLOV A N.Prediction of complex oscillations in the dynamics of coupled chaotic systems using transients [J].Physica A:Statistical Mechanics and Its Applications,2020,545:123818.
参考文献 5
VÜLLINGS A,HIZANIDIS J,OMELCHENKO I,et al.Clustered chimera states in systems of type-I excitability [J].New Journal of Physics,2014,16(12):123039.
参考文献 6
SCHMIDT L,SCHÖNLEBER K,KRISCHER K,et al.Coexistence of synchrony and incoherence in oscillatory media under nonlinear global coupling [J].Chaos:An Interdisciplinary Journal of Nonlinear Science,2014,24(1):013102.
参考文献 7
TAMAKI M,BANG J W,WATANABE T,et al.Night watch in one brain hemisphere during sleep associated with the first-night effect in humans [J].Current Biology,2016,26(9):1190-1194.
参考文献 8
UHLHAAS P J,SINGER W.Neural synchrony in brain disorders:relevance for cognitive dysfunctions and pathophysiology [J].Neuron,2006,52(1):155-168.
参考文献 9
FAN D G,WANG Q Y.Synchronization and bursting transition of the coupled Hindmarsh-Rose systems with asymmetrical time-delays [J].Science China Technological Science,2017,60(7):1019-1031.
参考文献 10
YANG Y,XIANG C,DAI X,et al.Chimera states and cluster solutions in Hindmarsh-Rose neural networks with state resetting process [J].Cognitive Neurodynamics,2022,16(1):215-228.
参考文献 11
ZHANG Y,WANG P,ZHANG M,et al.Chimera dynamics in nonlocally coupled bicomponent oscillators [J].Europhysics Letters,2021,135:40004.
参考文献 12
CALIM A,TORRES J J,OZER M,et al.Chimera states in hybrid coupled neuron populations [J].Neural Networks,2020,126:108-117.
参考文献 13
HUO S,TIAN C,KANG L,et al.Chimera states of neuron networks with adaptive coupling [J].Nonlinear Dynamics,2019,96(1):75-86.
参考文献 14
HIZANIDIS J,KOUVARIS N E,ZAMORA-LóPEZ G,et al.Chimera-like states in modular neural networks [J].Scientific Reports,2016,6(1):1-11.
参考文献 15
BERA B K,GHOSH D,BANERJEE T.Imperfect traveling chimera states induced by local synaptic gradient coupling [J].Physical Review E,2016,94(1):012215.
参考文献 16
SHEPELEV I A,BUKH A V,STRELKOVA G I,et al.Chimera states in ensembles of bistable elements with regular and chaotic dynamics [J].Nonlinear Dynamics,2017,90(4):2317-2330.
参考文献 17
KASATKIN D V,YANCHUK S,SCHLL E,et al.Self-organized emergence of multilayer structure and chimera states in dynamical networks with adaptive couplings [J].Physical Review E,2017,96(6):062211.
参考文献 18
ALBERT R,BARABÁSI A L.Statistical mechanics of complex networks [J].Reviews of Modern Physics,2002,74(1):47.
参考文献 19
NEWMAN M E J.The structure and function of complex networks [J].SIAM Review,2003,45(2):167-256.
参考文献 20
WANG D L.Emergent synchrony in locally coupled neural oscillators [J].IEEE Transactions on Neural Networks,1995,6(4):941-948.
参考文献 21
BERA B K,GHOSH D.Chimera states in purely local delay-coupled oscillators [J].Physical Review E,2016,93(5):052223.
参考文献 22
RUZZENE G,OMELCHENKO I,SAWICKI J,et al.Remote pacemaker control of chimera states in multilayer networks of neurons [J].Physical Review E,2020,102(5):052216.
参考文献 23
MAJHI S,PERC M,GHOSH D.Chimera states in uncoupled neurons induced by a multilayer structure [J].Scientific Reports,2016,6:39033.
参考文献 24
王松,茅晓晨.含时滞的忆阻耦合HR神经元的复杂放电行为 [J].动力学与控制学报,2020,18(1):33-39.WANG S,MAO X C.Complex firing behaviors of a memristive HR neurons with time delays [J].Journal of Dynamics and Control,2020,18(1):33-39.(in Chinese)
参考文献 25
马杰,高洁,独盟盟等.时滞下忆阻突触耦合Hopfield神经网络的动力学行为分析 [J].动力学与控制学报,2021,19(6):59-66.MA J,GAO J,DU M M,et al.Dynamic behavior analysis of memristive synaptic coupling Hpofiled neural network under time delay [J].Journal of Dynamics and Control,2021,19(6):59-66.(in Chinese)
参考文献 26
LIU Y,SUN Z,YANG X,et al.Dynamical robustness and firing modes in multilayer memristive neural networks of nonidentical neurons [J].Applied Mathematics and Computation,2021,409:126384.
参考文献 27
XU F,ZHANG J,JIN M,et al.Chimera states and synchronization behavior in multilayer memristive neural networks [J].Nonlinear Dynamics,2018,94(2):1-9.
参考文献 28
GOPAL R,CHANDRASEKAR V K,VENKATESAN A,et al.Observation and characterization of chimera states in coupled dynamical systems with nonlocal coupling [J].Physical Review E,2014,89(5):052914.
参考文献 29
OMELCHENKO I,OMELCHENKO E,HÖVEL P,et al.When nonlocal coupling between oscillators becomes stronger:patched synchrony ormultichimera states [J].Physical Review Letter,2013,110:224101.
参考文献 30
OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.
参考文献 31
RAKSHIT S,MAJHI S,BERA B K,et al.Time-varying multiplex network:Intralayer and interlayer synchronization [J].Physical Review E,2017,96(6):062308.
参考文献 32
YUAN Y Y,YANG H,HAN F,et al.Traveling chimera states in locally coupled memristive Hindmarsh-Rose neuronal networks and circuit simulation [J].Science China Technological Sciences,2022,65(7):1445-1455.
目录contents

    摘要

    嵌合态被发现存在于神经系统并且可能在神经元节律、大脑的睡眠和记忆等诸多神经过程中发挥重要作用.本文考虑神经元交互中的电磁感应现象,建立了以Hindmarsh-Rose神经元为节点的局部耦合双层忆阻神经元网络,研究其嵌合态时空动力学模式及产生机理.结果表明,改变层内、层间突触耦合强度会使网络产生移动和不完美移动嵌合态等多种类型的嵌合模式,其中不完美移动嵌合态中不相干的区域会扩展到网络的相干域.特别地,在特定耦合强度下,存在一种新的嵌合态活动模式,即一部分神经元处于嵌合态,另一部分神经元处于移动嵌合态.考虑神经元突触的忆阻特性,发现忆阻参数的增加能够使处于嵌合态的神经元网络转变为同步态,且耦合强度越大,达到同步态所需要的忆阻参数值越小.进一步探究双层网络的同步性,发现层间耦合强度和忆阻参数的增大有助于网络达到更好的同步.研究结果表明神经元之间的相互作用可以激发双层神经元网络产生多种嵌合态模式,电磁感应可以促进网络由嵌合态向同步态转迁,这些结果有助于理解人脑中复杂的神经放电过程和信息处理机制,并为可能的类脑装置应用提供参考.

    Abstract

    Chimera states have been found in nervous systems and may play an important role in many neural processes including neuronal rhythms, sleep and memory in the brain. Considering the phenomenon of electromagnetic induction in the neuron interactions, a two-layer memristive neuronal network of locally coupled Hindmarsh-Rose neurons was established to study the spatiotemporal dynamical patterns of chimera states and the mechanisms of their generation. It is found that changing the intra- and inter-layer synaptic coupling strengths can induce several types of chimera states in networks, including traveling and imperfect traveling chimera states where the incoherent traveling domain spreads into the coherent domain of the network. In particular, a phenomenon that several neurons are in chimera states and others are in traveling chimera states is witnessed at specific coupling strengths. Moreover, it is also observed that increasing the memristive parameter can induce the neuronal network to change from chimera states to synchronous states, and the stronger the coupling strengths, the smaller the value of the memristive parameter required to reach synchronous states. Considering the inter-layer synchronization, the increase of the coupling strengths and the memristive parameter helps the network to achieve better synchronisation. The results demonstrate that the interactions between neurons on two-layer neuronal networks can stimulate a variety of patterns of chimera states, and that the effect of electromagnetic induction can facilitate the transition from chimera states to coherent states. These results are useful for understanding the complex neural firing processes and information processing mechanisms in the human brain and may provide a reference for the application of the brain-like devices.

  • 引言

  • 人脑中有大量神经元,这些神经元组成复杂的网络结构参与神经信息处理以及大脑的其他活动,通过建模研究神经元网络的动态活动机制具有十分重要的意义.2002年,Kuramoto在由相同振荡器通过非局部耦合形成的系统中发现了一种同步与非同步态共存的复杂动态模式[1],后来Abrams等人[2]引用古希腊神话中的神兽将这种奇特的现象命名为“Chimera”.二十年来,人们对嵌合态的研究逐渐深入,在生物、物理和化学等多学科领域进行了广泛探索,证实它不仅可以存在于相位振荡器,还存在于混沌系统[34]和Hopf范式[56]等诸多系统中.在神经系统中,自然界中的多种神经过程都证明了嵌合态的存在,如一些鸟类和水生哺乳动物的单半球慢波睡眠现象和人类的首晚现象[7],这种同步与不同步共存的现象还与帕金森症、癫痫发作等多种神经疾病密切相关[89].因此,神经元网络中的嵌合态获得了广泛的关注,在Hindmarsh-Rose[910]、FitzHugh-Nagumo[11]和Morris-Lecar[12]等多种神经元模型构成的多种网络拓扑结构中都观察到了嵌合态,Huo等人[13]研究了随机网络、全连接网络和无标度网络中的嵌合态特征,证实了网络结构会影响嵌合态出现的可能性.Hizanidis等人[14]建立了基于秀丽隐杆线虫土壤蠕虫连接组的神经元网络模型,证实了模块化网络也会出现类嵌合态现象.为了区分嵌合态的多种特征,人们根据时空动力学将其划分为移动嵌合态、暂态嵌合态、不完美嵌合态[1516]等类型,同时对可能影响嵌合态存在的耦合强度、网络拓扑和时延[17]等多种因素进行了深入研究.

  • 近些年来,对复杂网络集体行为的研究在生物科学、计算机技术和统计物理等[1819]多个领域引起了广泛重视.局部耦合网络是常见的规则网络连接方式,它在感知分组和模式分割中有很大潜力[20],且局部耦合的非线性特征能在嵌合态的产生中发挥重要作用[21].大脑皮层网络十分复杂,单层网络不足以说明系统功能,需要用多层网络结构模拟更加真实的系统.多层结构势必会对神经元网络的动力学行为产生影响.特别地,在多层网络中也发现了嵌合态现象,Ruzzene等人[22]通过构建由FitzHugh-Nagumo神经元组成的多层网络,研究了利用复用结构远程控制嵌合态的可能性,为控制嵌合态提供了新思路.Majhi等人[23]研究了多层神经元网络中突触和信息传输延迟对嵌合态和同步簇模式的影响.

  • 神经元之间的信息传递主要通过电突触和化学突触两种功能性连接进行,已经在电和化学突触耦合的多种神经元网络中观察到嵌合态现象.生物神经网络中信号传输和离子交换可能会产生复杂的电磁场,电磁场与膜电位之间存在忆阻联系,并且神经元突触连接本质上是有记忆特性的.因此选择忆阻突触构建神经元网络能够有效模拟生物突触动态行为,也可以帮助解释神经系统的信息处理和调控机制.王松等[24]研究含有忆阻突触的两耦合神经元系统,证实了信号传输中的时滞可以影响系统稳定性和放电模式.马杰等[25]研究了含有时滞的忆阻突触耦合的四维Hopfield神经网络,发现时滞和忆阻耦合强度会使系统产生多种时空动力学行为.Liu等[26]构建由不同神经元组成的层间忆阻突触连接的多层神经元网络,探究其动态鲁棒性及放电模式的转换,发现弱电磁感应会破坏动态鲁棒性.但是在生物神经元网络中突触的记忆性广泛存在,探究神经元之间的忆阻特性与复杂网络的时空动力学特征之间的关系具有十分重要的意义.

  • 本文着力于探究忆阻突触耦合的神经元网络中的嵌合态时空动力学行为,构建了一个基于Hindmarsh-Rose神经元的局部耦合双层神经元网络,层内和层间节点都用忆阻突触连接.首先,研究在不同耦合强度下神经元网络出现的多种嵌合态动力学行为.然后,探讨忆阻参数对神经元网络动力学行为的影响和不同参数与网络层间同步性的关系.

  • 1 模型描述与分析方法

  • 1.1 神经元网络模型

  • 本文中,神经元网络为双层结构,单个振子与其在层内两个方向上最近邻和另一层网络对应位置上的神经元连接(称为局部耦合),用Hindmarsh-Rose神经元作为网络的节点,用忆阻突触模拟生物突触连接层内和层间神经元.网络结构如图1所示(红色和紫色的圆分别代表网络第一层和第二层中的神经元,蓝色的长方形代表忆阻突触),网络动力学用如下方程描述:

  • 图1 局部耦合的双层忆阻神经元网络结构图

  • Fig.1 Schematic diagram of the two-layer memristive neural network of locally coupled

  • x˙i, 1=axi, 12-xi, 13-yi, 1-zi, 1+

  • ε1Mφi-1, 1xi-1, 1-xi, 1+Mφi, 1xi+1, 1-

  • xi, 1-λ3Mφi, 12xi, 1-xi, 2

  • y˙i, 1= (a+α) xi, 12-yi, 1

  • z˙i, 1=wbxi, 1-zi, 1+c

  • φ˙i, 1=xi, 1-xi+1, 1-δ1φi, 1

  • x˙i, 2=axi, 22-xi, 23-yi, 2-zi, 2+

  • ε2Mφi-1, 2xi-1, 2-xi, 2+Mφi, 2xi+1, 2-

  • xi, 2+λ3Mφi, 12xi, 1-xi, 2

  • y˙i, 2= (a+α) xi, 22-yi, 2

  • z˙i, 2=wbxi, 2-zi, 2+c

  • φ˙i, 2=xi, 2-xi+1, 2-δ2φi, 2

  • φ˙i,12=xi,1-xi,2-δ3φi,12
    (1)
  • 式中,xyz分别表示单个神经元的膜电位、恢复变量和与钠、钾离子有关的慢适应电流.N表示每层网络中的神经元数量,x0j=xNjxN+1j=x1jj=1,2表示两层结构,式中的δ表示忆阻器的遗忘参数,参数ε代表每层神经元之间忆阻突触的耦合强度,λ3代表层间忆阻突触的耦合强度,变量φi,jMφi,j)分别表示第j行第i个忆阻突触的磁通量和忆感,Mφi,j)是qφi,j)电荷相对于磁通量φi,j的导数,使用三阶磁通控制忆阻突触,其电荷方程表示为:

  • qφi,j=σφi,j+θφi,j3
    (2)
  • 对磁通量求导得到忆阻的忆感表示为:

  • Mφi,j=σ+3θφi,j2
    (3)
  • 式中,σθ是忆阻器的参数,它们的取值与环境和忆阻器自身条件有关,将神经元的参数设置为a=1.45,α=1.6,u=0.001,b=9,c=5,δj=0.5,δ3=0.5,θ=0.02.

  • 对式(1)中每个神经元的变量设置初始值,参考文献[27],设置前N/2个神经元为有规律的初始值,后N/2个神经元的初始条件加入小的随机波动,表示为:当i=1,2,···,N/2时,xi0,j=0.01(i-N/2),yi0,j=0.02(i-N/2),zi0,j=0.03(i-N/2),对其他神经元设置为xi0,j=0.1(N/2-i),yi0,j=0.12(N/2-i),zi0,j=0.21(N/2-i).

  • 1.2 分析方法

  • 为了更好地区分嵌合态与其他状态,引入不相干强度(SI,the strength of incoherence)[28]这一统计变量,首先对相邻神经元膜电位xi,j求取差值,得到wij=xij-xi+1jwij=w1ijw2ijwdijRd,为了更加清楚地量化同步状态,将每层神经元分为M组大小为n=N/M的部分,因此局部标准变量(local standard deviation)ηlm)可以描述为:

  • ηi,j(m)=1nk=n(m-1)+1mn wl,k,j-wi,jt
    (4)
  • m=1,2,···,Ml=1,2,···,di=1,2,···,Nd为神经元模型的维度,wij=1ni=1N wlijtt代表一段时间内的平均值,ηi,jm)可以计算每n个连续神经元的局部标准偏差,不相干强度描述为:

  • SIj=1-m=1M sm,jM,sm,j=Θδ0-ηi,j(m)
    (5)
  • 其中Θ(·)是阶跃函数,δ0是提前设定的阈值并通常很小,SI=0,SI=1和0<SI<1分别表示同步,不同步和嵌合态.

  • 另外,使用局部序参数(local order parameter)的统计量[29]更加直观地辨别嵌合态、不同步态和同步态,局部序参数代表神经元网络的局部有序性,表示不相干的程度.其公式描述如下:

  • Li,j=12p|i-k|p eηΦk
    (6)
  • 式中η=-1i=1,2N表示神经元个数,k=1,2,···,Nj=1,2.参数p指每个神经元两边最近邻的数量,Φk=arctan(yk/xk)是第k个神经元的几何相位[30].Li,j≈0表明第j层第i个神经元属于不相干态的节点,Li,j≈1表明第j层第i个神经元属于相干部分,Li,j=1代表完全同步.

  • 为了进一步探讨两层网络动态行为之间的关系,引入层间同步误差[31],其公式描述为:

  • E=limΔT 1ΔTΔT 1(M-1)Nj=1M-1 i=1N xi,j-xi,j+1dt
    (7)
  • MN分别代表网络中的层数和每一层神经元的数量.E=0代表网络达到最佳同步状态.

  • 2 结果与分析

  • 2.1 耦合强度诱发多种嵌合态

  • 首先,将层内和层间突触耦合强度参数设置为ε1=5.5、λ3=0.1、ε2=0,第一层和第二层层内耦合强度较大的差值使网络的两层呈现出不同的状态,此时忆阻参数固定在σ=0.12[32].图1给出了双层网络时空响应图、神经元时间序列、神经元群在某一时刻的膜电位分布和局部序参量时空图.从图2(b)可以看出第一层网络中一些神经元同步,而其余的神经元不同步,这是嵌合态的特征.图2(a)显示了第一层所有神经元的膜电位在一段时间内随时间演化过程,神经元的同步状态随时间发生变化,呈现出移动嵌合态,图2(c)中的局部序参量的时空响应也与之相符.第二层网络的神经元在移动过程中一些相干态的神经元被转化为不相干态,这种非相干域扩展到相干域的特殊移动嵌合态是不完美移动嵌合态.图2(d)清楚地显示了在时间和空间移动的过程中宽度变化的时空图,从时刻所有神经元膜电位分布的快照可以观察到相干与不相干的神经元共存,有三组不相干簇,放电方式为高原簇放电,如图2(e)所示.计算出每个神经元的局部序参量,如图2(f)所示,黄色的区域表示相干节点,两个连续的相干域之间蓝色的不相干域在时空中传播时,其宽度会发生变化,出现不相干域向相干域扩展的情况,这是不完美移动嵌合态的明确标志.

  • 图2 在ε1=5.5、λ3=0.1、ε2=0下双层忆阻神经网络的时空响应图、膜电位分布情况及时间序列、局部序参量

  • Fig.2 Spatiotemporal plots, time series and the distribution, the local order parameter in two-layermemristive neuronal network with ε1=5.5, λ3=0.1 and ε2=0

  • 特别地,当层间耦合强度较大但层内耦合强度都很小,即ε1=0.02、λ3=5、ε2=0.02时,两层网络都出现了嵌合态与移动嵌合态共存的现象.以第一层网络为例,神经元膜电位时空分布图呈现出规则的条纹状,而右半部分为倾斜的条纹,表示神经元的状态随着时间推移发生变化,如图3(a)所示.图3(b)表示t=3900ms时所有神经元的膜电位分布情况,以i=40为界分为两部分,第Ⅰ部分神经元处于无序嵌合态,第Ⅱ部分的神经元被划分为几个相干簇,一部分神经元放电,另一部分属于静息,少数神经元处于不相干态,且每个神经元的同步状态会随着时间的推移发生改变,处于移动嵌合态.局部序参量的时空响应图更加清楚地表现这一现象,如图3(c)所示,第Ⅰ部分中相干域与不相干域交替出现且保持静态,属于嵌合态,第Ⅱ部分中神经元相干与不相干的状态随时间发生变化,属于移动嵌合态.由此得出,耦合强度对多种类型嵌合态的产生有重要作用.

  • 图3 在ε1=0.02、λ3=5、ε2=0.02下忆阻神经网络第一层的时空响应图、膜电位分布情况及时间序列、局部序参量

  • Fig.3 Spatiotemporal plots, time series and the distribution, the local order parameter in the first layer of the memristive neuronal network with ε1=0.02, λ3=5 and ε2=0.02

  • 2.2 忆阻参数促进嵌合态向同步转迁

  • 调整层内和层间耦合强度,神经元网络中没有出现同步态现象.当ε1=1.5、λ3=0.8、ε2=0时,神经元网络表现为移动嵌合态,此时忆阻参数σ=0.12.在耦合强度不变的条件下增大忆阻参数,到σ=4.2时所有神经元的膜电位显示出周期放电的模式,且神经元间膜电位彼此关联,如图4(a)、4(b)所示,神经元处于完全同步态.图4(c)计算出一段时间内的局部序参量,所有神经元在时空内都保持同步状态.由此看出,忆阻参数σ的增大能打破神经元网络的嵌合态,最终使所有神经元均处于同步态.

  • 采用不相干强度SI来定量分析双层网络的动力学行为,图5(a)、5(b)分别是两层网络中设置λ3=0.8、ε2=0而ε1取1.5、2、3不同值时,所有神经元的状态随忆阻参数变化的情况.可以看出不同层间耦合强度值ε1=1.5、2、3,网络先处于嵌合态(0<SI<1),忆阻参数增大到σ=4.5、3.5、3之后网络呈现同步态(SI=0),且层内耦合越大,想要达到同步态所需要的忆阻参数值越小.同样地,图5(c)和5(d)分别表示两层网络在不同强度层间作用时,神经元状态随忆阻参数变化的情况,此时ε1=3、ε2=0.随着层间作用的增强,神经元网络可以在较小的忆阻参数时就从嵌合态转变为同步态.由此可以得出,层内、层间相互作用和忆阻参数的增加都可促进网络向同步态的转迁.

  • 图4 在ε1=1.5、λ3=0.8、ε2=0、σ=4.2下忆阻神经网络第一层的时空响应图、膜电位分布情况及时间序列、局部序参量

  • Fig.4 Spatiotemporal plots, time series and the distribution, the local order parameter in the first layer of the memristive neuronal network with ε1=1.5, λ3=0.8, ε2=0, σ=4.2

  • 图5 双层神经元网络中不同耦合强度下不相干强度随忆阻参数变化曲线

  • Fig.5 Variation of the strength of incoherence with the memristive parameter for different values of coupling strengths in the two-layer neuronal network

  • 2.3 不同参数对层间同步性的影响

  • 为了探究层内、层间相互作用和忆阻突触对层间同步动态行为的影响,设置、,作出神经元网络在不同忆阻参数下层间同步误差随层间耦合强度变化情况的曲线图.从图6(a)可以看出,随着忆阻参数的增大层间同步误差有微小的下降,没有层间作用时层间误差很大,一旦层间作用增大,误差就迅速减小,并且层间耦合强度持续增大层间误差一直保持在较小的值.图6(b)是ε1=1、σ=0.12且层间耦合强度取不同值时,层间误差随层内耦合变化的情况,ε2≤1时层间误差随层内耦合的增大而减小,在ε2=1时两层网络层内耦合强度相同,每个神经元受到来自其他神经元的输入电流相同,误差为零,网络达到完全同步状态,而ε2>1时误差随层内耦合的增大而增大,但是在层间作用较大时这种变化并不明显.综上,层间作用和忆阻参数的增大均会促进双层网络的层间同步.

  • 图6 层间同步误差变化曲线

  • Fig.6 Variation curve of inter-layer synchronization error

  • 3 结论

  • 本文构建了一个局部耦合的忆阻突触连接的Hindmarsh-Rose双层神经元网络,研究不同耦合强度和忆阻参数下神经元网络的嵌合态动力学.首先调节层内和层间耦合强度,出现两层网络分别表现不同状态的情况,一层网络中的神经元状态为随时间发生变化的移动嵌合态,另一层出现移动过程中非相干域扩展到相干域的现象,为不完美移动嵌合态.继续改变耦合强度值,发现两层都出现了一部分神经元处于嵌合态,另一部分神经元处于移动嵌合态的现象,说明耦合强度可以激发神经元网络出现多种嵌合态行为.然后,增大忆阻突触参数,观察到了调整耦合强度时没有出现的同步态,说明忆阻参数的增加能促使网络由嵌合态转迁为同步态.进一步发现层内和层间耦合强度增大,忆阻参数在更小值就能使网络转变为同步态.最后探究了不同参数对层间同步性的影响,结果表明增大层间耦合强度和忆阻参数会促进双层网络层间同步,且在两层网络层内耦合强度相同时达到最佳同步状态.上述研究对探索神经系统中的信息处理和放电节律有关机制有重要帮助,可能为类脑装置的实现提供了新思路.

  • 参考文献

    • [1] KURAMOTO Y,BATTOGTOKH D.Coexistence of coherence and incoherence in nonlocally coupled phase oscillators [J].Nonlinear Phenomena in Complex Systems,2002,5(4):380-385.

    • [2] ABRAMS D M,STROGATZ S H.Chimera states for coupled oscillators [J].Physical Review Letters,2004,93(17):174102.

    • [3] OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.

    • [4] PAVLOVA O N,PAVLOV A N.Prediction of complex oscillations in the dynamics of coupled chaotic systems using transients [J].Physica A:Statistical Mechanics and Its Applications,2020,545:123818.

    • [5] VÜLLINGS A,HIZANIDIS J,OMELCHENKO I,et al.Clustered chimera states in systems of type-I excitability [J].New Journal of Physics,2014,16(12):123039.

    • [6] SCHMIDT L,SCHÖNLEBER K,KRISCHER K,et al.Coexistence of synchrony and incoherence in oscillatory media under nonlinear global coupling [J].Chaos:An Interdisciplinary Journal of Nonlinear Science,2014,24(1):013102.

    • [7] TAMAKI M,BANG J W,WATANABE T,et al.Night watch in one brain hemisphere during sleep associated with the first-night effect in humans [J].Current Biology,2016,26(9):1190-1194.

    • [8] UHLHAAS P J,SINGER W.Neural synchrony in brain disorders:relevance for cognitive dysfunctions and pathophysiology [J].Neuron,2006,52(1):155-168.

    • [9] FAN D G,WANG Q Y.Synchronization and bursting transition of the coupled Hindmarsh-Rose systems with asymmetrical time-delays [J].Science China Technological Science,2017,60(7):1019-1031.

    • [10] YANG Y,XIANG C,DAI X,et al.Chimera states and cluster solutions in Hindmarsh-Rose neural networks with state resetting process [J].Cognitive Neurodynamics,2022,16(1):215-228.

    • [11] ZHANG Y,WANG P,ZHANG M,et al.Chimera dynamics in nonlocally coupled bicomponent oscillators [J].Europhysics Letters,2021,135:40004.

    • [12] CALIM A,TORRES J J,OZER M,et al.Chimera states in hybrid coupled neuron populations [J].Neural Networks,2020,126:108-117.

    • [13] HUO S,TIAN C,KANG L,et al.Chimera states of neuron networks with adaptive coupling [J].Nonlinear Dynamics,2019,96(1):75-86.

    • [14] HIZANIDIS J,KOUVARIS N E,ZAMORA-LóPEZ G,et al.Chimera-like states in modular neural networks [J].Scientific Reports,2016,6(1):1-11.

    • [15] BERA B K,GHOSH D,BANERJEE T.Imperfect traveling chimera states induced by local synaptic gradient coupling [J].Physical Review E,2016,94(1):012215.

    • [16] SHEPELEV I A,BUKH A V,STRELKOVA G I,et al.Chimera states in ensembles of bistable elements with regular and chaotic dynamics [J].Nonlinear Dynamics,2017,90(4):2317-2330.

    • [17] KASATKIN D V,YANCHUK S,SCHLL E,et al.Self-organized emergence of multilayer structure and chimera states in dynamical networks with adaptive couplings [J].Physical Review E,2017,96(6):062211.

    • [18] ALBERT R,BARABÁSI A L.Statistical mechanics of complex networks [J].Reviews of Modern Physics,2002,74(1):47.

    • [19] NEWMAN M E J.The structure and function of complex networks [J].SIAM Review,2003,45(2):167-256.

    • [20] WANG D L.Emergent synchrony in locally coupled neural oscillators [J].IEEE Transactions on Neural Networks,1995,6(4):941-948.

    • [21] BERA B K,GHOSH D.Chimera states in purely local delay-coupled oscillators [J].Physical Review E,2016,93(5):052223.

    • [22] RUZZENE G,OMELCHENKO I,SAWICKI J,et al.Remote pacemaker control of chimera states in multilayer networks of neurons [J].Physical Review E,2020,102(5):052216.

    • [23] MAJHI S,PERC M,GHOSH D.Chimera states in uncoupled neurons induced by a multilayer structure [J].Scientific Reports,2016,6:39033.

    • [24] 王松,茅晓晨.含时滞的忆阻耦合HR神经元的复杂放电行为 [J].动力学与控制学报,2020,18(1):33-39.WANG S,MAO X C.Complex firing behaviors of a memristive HR neurons with time delays [J].Journal of Dynamics and Control,2020,18(1):33-39.(in Chinese)

    • [25] 马杰,高洁,独盟盟等.时滞下忆阻突触耦合Hopfield神经网络的动力学行为分析 [J].动力学与控制学报,2021,19(6):59-66.MA J,GAO J,DU M M,et al.Dynamic behavior analysis of memristive synaptic coupling Hpofiled neural network under time delay [J].Journal of Dynamics and Control,2021,19(6):59-66.(in Chinese)

    • [26] LIU Y,SUN Z,YANG X,et al.Dynamical robustness and firing modes in multilayer memristive neural networks of nonidentical neurons [J].Applied Mathematics and Computation,2021,409:126384.

    • [27] XU F,ZHANG J,JIN M,et al.Chimera states and synchronization behavior in multilayer memristive neural networks [J].Nonlinear Dynamics,2018,94(2):1-9.

    • [28] GOPAL R,CHANDRASEKAR V K,VENKATESAN A,et al.Observation and characterization of chimera states in coupled dynamical systems with nonlocal coupling [J].Physical Review E,2014,89(5):052914.

    • [29] OMELCHENKO I,OMELCHENKO E,HÖVEL P,et al.When nonlocal coupling between oscillators becomes stronger:patched synchrony ormultichimera states [J].Physical Review Letter,2013,110:224101.

    • [30] OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.

    • [31] RAKSHIT S,MAJHI S,BERA B K,et al.Time-varying multiplex network:Intralayer and interlayer synchronization [J].Physical Review E,2017,96(6):062308.

    • [32] YUAN Y Y,YANG H,HAN F,et al.Traveling chimera states in locally coupled memristive Hindmarsh-Rose neuronal networks and circuit simulation [J].Science China Technological Sciences,2022,65(7):1445-1455.

  • 参考文献

    • [1] KURAMOTO Y,BATTOGTOKH D.Coexistence of coherence and incoherence in nonlocally coupled phase oscillators [J].Nonlinear Phenomena in Complex Systems,2002,5(4):380-385.

    • [2] ABRAMS D M,STROGATZ S H.Chimera states for coupled oscillators [J].Physical Review Letters,2004,93(17):174102.

    • [3] OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.

    • [4] PAVLOVA O N,PAVLOV A N.Prediction of complex oscillations in the dynamics of coupled chaotic systems using transients [J].Physica A:Statistical Mechanics and Its Applications,2020,545:123818.

    • [5] VÜLLINGS A,HIZANIDIS J,OMELCHENKO I,et al.Clustered chimera states in systems of type-I excitability [J].New Journal of Physics,2014,16(12):123039.

    • [6] SCHMIDT L,SCHÖNLEBER K,KRISCHER K,et al.Coexistence of synchrony and incoherence in oscillatory media under nonlinear global coupling [J].Chaos:An Interdisciplinary Journal of Nonlinear Science,2014,24(1):013102.

    • [7] TAMAKI M,BANG J W,WATANABE T,et al.Night watch in one brain hemisphere during sleep associated with the first-night effect in humans [J].Current Biology,2016,26(9):1190-1194.

    • [8] UHLHAAS P J,SINGER W.Neural synchrony in brain disorders:relevance for cognitive dysfunctions and pathophysiology [J].Neuron,2006,52(1):155-168.

    • [9] FAN D G,WANG Q Y.Synchronization and bursting transition of the coupled Hindmarsh-Rose systems with asymmetrical time-delays [J].Science China Technological Science,2017,60(7):1019-1031.

    • [10] YANG Y,XIANG C,DAI X,et al.Chimera states and cluster solutions in Hindmarsh-Rose neural networks with state resetting process [J].Cognitive Neurodynamics,2022,16(1):215-228.

    • [11] ZHANG Y,WANG P,ZHANG M,et al.Chimera dynamics in nonlocally coupled bicomponent oscillators [J].Europhysics Letters,2021,135:40004.

    • [12] CALIM A,TORRES J J,OZER M,et al.Chimera states in hybrid coupled neuron populations [J].Neural Networks,2020,126:108-117.

    • [13] HUO S,TIAN C,KANG L,et al.Chimera states of neuron networks with adaptive coupling [J].Nonlinear Dynamics,2019,96(1):75-86.

    • [14] HIZANIDIS J,KOUVARIS N E,ZAMORA-LóPEZ G,et al.Chimera-like states in modular neural networks [J].Scientific Reports,2016,6(1):1-11.

    • [15] BERA B K,GHOSH D,BANERJEE T.Imperfect traveling chimera states induced by local synaptic gradient coupling [J].Physical Review E,2016,94(1):012215.

    • [16] SHEPELEV I A,BUKH A V,STRELKOVA G I,et al.Chimera states in ensembles of bistable elements with regular and chaotic dynamics [J].Nonlinear Dynamics,2017,90(4):2317-2330.

    • [17] KASATKIN D V,YANCHUK S,SCHLL E,et al.Self-organized emergence of multilayer structure and chimera states in dynamical networks with adaptive couplings [J].Physical Review E,2017,96(6):062211.

    • [18] ALBERT R,BARABÁSI A L.Statistical mechanics of complex networks [J].Reviews of Modern Physics,2002,74(1):47.

    • [19] NEWMAN M E J.The structure and function of complex networks [J].SIAM Review,2003,45(2):167-256.

    • [20] WANG D L.Emergent synchrony in locally coupled neural oscillators [J].IEEE Transactions on Neural Networks,1995,6(4):941-948.

    • [21] BERA B K,GHOSH D.Chimera states in purely local delay-coupled oscillators [J].Physical Review E,2016,93(5):052223.

    • [22] RUZZENE G,OMELCHENKO I,SAWICKI J,et al.Remote pacemaker control of chimera states in multilayer networks of neurons [J].Physical Review E,2020,102(5):052216.

    • [23] MAJHI S,PERC M,GHOSH D.Chimera states in uncoupled neurons induced by a multilayer structure [J].Scientific Reports,2016,6:39033.

    • [24] 王松,茅晓晨.含时滞的忆阻耦合HR神经元的复杂放电行为 [J].动力学与控制学报,2020,18(1):33-39.WANG S,MAO X C.Complex firing behaviors of a memristive HR neurons with time delays [J].Journal of Dynamics and Control,2020,18(1):33-39.(in Chinese)

    • [25] 马杰,高洁,独盟盟等.时滞下忆阻突触耦合Hopfield神经网络的动力学行为分析 [J].动力学与控制学报,2021,19(6):59-66.MA J,GAO J,DU M M,et al.Dynamic behavior analysis of memristive synaptic coupling Hpofiled neural network under time delay [J].Journal of Dynamics and Control,2021,19(6):59-66.(in Chinese)

    • [26] LIU Y,SUN Z,YANG X,et al.Dynamical robustness and firing modes in multilayer memristive neural networks of nonidentical neurons [J].Applied Mathematics and Computation,2021,409:126384.

    • [27] XU F,ZHANG J,JIN M,et al.Chimera states and synchronization behavior in multilayer memristive neural networks [J].Nonlinear Dynamics,2018,94(2):1-9.

    • [28] GOPAL R,CHANDRASEKAR V K,VENKATESAN A,et al.Observation and characterization of chimera states in coupled dynamical systems with nonlocal coupling [J].Physical Review E,2014,89(5):052914.

    • [29] OMELCHENKO I,OMELCHENKO E,HÖVEL P,et al.When nonlocal coupling between oscillators becomes stronger:patched synchrony ormultichimera states [J].Physical Review Letter,2013,110:224101.

    • [30] OMELCHENKO I,MAISTRENKO Y,HÖVEL P,et al.Loss of coherence in dynamical networks:spatial chaos and chimera states [J].Physical Review Letters,2011,106(23):234102.

    • [31] RAKSHIT S,MAJHI S,BERA B K,et al.Time-varying multiplex network:Intralayer and interlayer synchronization [J].Physical Review E,2017,96(6):062308.

    • [32] YUAN Y Y,YANG H,HAN F,et al.Traveling chimera states in locally coupled memristive Hindmarsh-Rose neuronal networks and circuit simulation [J].Science China Technological Sciences,2022,65(7):1445-1455.

  • 微信公众号二维码

    手机版网站二维码