搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

一种新型复合指数型局部有源忆阻器耦合的Hopfield神经网络

王梦蛟 杨琛 贺少波 李志军

引用本文:
Citation:

一种新型复合指数型局部有源忆阻器耦合的Hopfield神经网络

王梦蛟, 杨琛, 贺少波, 李志军

A novel compound exponential locally active memristor coupled Hopfield neural network

Wang Meng-Jiao, Yang Chen, He Shao-Bo, Li Zhi-Jun
PDF
HTML
导出引用
  • 由忆阻耦合的神经网络模型, 因其能更真实地反映生物神经系统的复杂动力学特性而被广泛研究. 目前用于耦合神经网络的忆阻器数学模型主要集中在一次函数、绝对值函数、双曲正切函数等, 为进一步丰富忆阻耦合神经网络模型, 且考虑到一些掺杂半导体中粒子的运动规律, 设计了一种新的复合指数型局部有源忆阻器, 并将其作为耦合突触用于Hopfield神经网络, 利用基本的动力学分析方法, 研究了系统在不同参数下的动力学行为, 以及在不同初始值下多种分岔模式共存的现象. 实验结果表明, 忆阻突触内部参数对系统具有调控作用, 且该系统拥有丰富的动力学行为, 包括对称吸引子共存、非对称吸引子共存、大范围的混沌状态和簇发振荡等. 最后, 用STM32单片机对系统进行了硬件实现.
    The neural network model coupled with memristors has been extensively studied due to its ability to more accurately represent the complex dynamic characteristics of the biological nervous system. Currently, the mathematical model of memristor used to couple neural networks mainly focuses on primary function, absolute value function, hyperbolic tangent function, etc. To further enrich the memristor-coupled neural network model and take into account the motion law of particles in some doped semiconductors, a new compound exponential local active memristor is proposed and used as a coupling synapse in the Hopfield neural network. Using the basic dynamic analysis method, the system’s dynamic behaviors are studied under different parameters and the coexistence of multiple bifurcation modes under different initial values. In addition, the influence of frequency change of external stimulation current on the system is also studied. The experimental results show that the internal parameters of memristor synapses regulate the system, and the system has a rich dynamic behavior, including symmetric attractor coexistence, asymmetric attractor coexistence, large-scale chaos as shown in attached figure, and bursting oscillation. Finally, the hardware of the system is realized by the STM32 microcontroller, and the experimental results verify the realization of the system.
      通信作者: 王梦蛟, wangmj@xtu.edu.cn
    • 基金项目: 国家自然科学基金(批准号: 62071411)资助的课题.
      Corresponding author: Wang Meng-Jiao, wangmj@xtu.edu.cn
    • Funds: Project supported by the National Natural Science Foundation of China (Grant No. 62071411).
    [1]

    Chua L 1971 IEEE Trans. Circuits Syst. I Regul. Pap. 18 507Google Scholar

    [2]

    Chua L 2014 Semicond. Sci. Technol. 29 104001Google Scholar

    [3]

    Wang M J, Deng Y, Liao X H, Li Z J, Ma M L, Zeng Y C 2019 Int. J. Non. Linear. Mech. 111 149Google Scholar

    [4]

    Peng Y X, Liu J, He S B, Sun K H 2023 Chaos, Solitons Fractals 171 113429Google Scholar

    [5]

    Wang M J, An M Y, He S B, Zhang X N, Iu H H, Li Z J 2023 Chaos 33 073129Google Scholar

    [6]

    Peng Y X, He S B, Sun K H 2022 Nonlinear Dyn. 107 1263Google Scholar

    [7]

    He S B, Liu J, Wang H H, Sun K H 2023 Neurocomputing (Amst. ) 523 1Google Scholar

    [8]

    古亚娜, 梁燕, 王光义, 夏晨阳 2022 物理学报 71 110501Google Scholar

    Gu Y N, Liang Y, Wang G Y, Xia C Y 2022 Acta Phys. Sin. 71 110501Google Scholar

    [9]

    武长春, 周莆钧, 王俊杰, 李国, 胡绍刚, 于奇, 刘洋 2022 物理学报 71 148401Google Scholar

    Wu C C, Zhou P J, Wang J J, Li G, Hu S G, Yu Q, Liu Y 2022 Acta Phys. Sin. 71 148401Google Scholar

    [10]

    Chua L 2013 Nanotechnology 24 383001Google Scholar

    [11]

    Lu Y C, Li H M, Li C L 2023 Neurocomputing (Amst. ) 544 126246Google Scholar

    [12]

    Li Z J, Chen K J 2023 Chaos, Solitons Fractals 175 114017Google Scholar

    [13]

    Hopfield J J 1984 Proc. Natl. Acad. Sci. U. S. A. 81 3088Google Scholar

    [14]

    Chen C J, Chen J Q, Bao H, Chen M, Bao B C 2019 Nonlinear Dyn. 95 3385Google Scholar

    [15]

    Lin H R, Wang C H, Hong Q H, Sun Y C 2020 IEEE Tran. Circuits Syst. II Express Briefs 67 3472Google Scholar

    [16]

    Chen C J, Min F H, Zhang Y Z, Bao B C 2021 Nonlinear Dyn. 106 2559Google Scholar

    [17]

    Li C L, Yang Y Y, Yang X B, Zi X Y, Xiao F L 2022 Nonlinear Dyn. 108 1697Google Scholar

    [18]

    Doubla I S, Ramakrishnan B, Njitacke Z T, Kengne J, Rajagopal K 2022 Int. J. Electron. Commun. 144 154059Google Scholar

    [19]

    黄丽丽, 黄强, 黄振, 臧红岩, 雷腾飞 2023 电子元件与材料 42 10Google Scholar

    Huang L L, Huang Q, Huang Z, Zang H Y, Lei T F 2023 Electron. Compon. Mater. 42 10Google Scholar

    [20]

    Lin H R, Wang C H, Sun J R, Zhang X, Sun Y C, Iu H H C 2023 Chaos Solitons Fractals 166 112905Google Scholar

    [21]

    Wan Q Z, Chen S M, Yang Q, Liu J, Sun K L 2023 Nonlinear Dyn. 111 18505Google Scholar

    [22]

    Lin H R, Wang C H, Yu F, Sun J R, Du S C, Deng Z K, Deng Q L 2023 Mathematics (Basel) 11 1369Google Scholar

    [23]

    Panahi S, Aram Z, Jafari S, Ma J, Sprott J C 2017 Chaos, Solitons Fractals 105 150Google Scholar

    [24]

    Guevara M R, Glass L, Mackey M C, Shrier A 1983 IEEE Trans. Syst. Man Cybern. Syst. 5 790Google Scholar

    [25]

    Chua L 2018 Appl. Phys. A: Mater. 124 563Google Scholar

    [26]

    Chua L 2005 Int. J. Bifurcat. Chaos 15 3435Google Scholar

    [27]

    Ascoli A, Slesazeck S, Mähne H, Tetzlaff R, Mikolajick T 2015 IEEE Trans. Circuits Syst. I Regul. Pap. 62 1165Google Scholar

    [28]

    Chua L 2015 Radioengineering 24 319Google Scholar

    [29]

    Wang M J, Li J H, Yu S S, Zhang X N, Li Z J, Iu H H C 2020 Chaos 30 043125Google Scholar

    [30]

    Bi Q S, Gou J T 2023 Chaos Solitons Fractals 167 113046Google Scholar

    [31]

    莱维坦 I B, 卡茨玛克 L K著 (舒斯云, 包新民 译) 2001 神经元: 细胞和分子生物学 (北京: 科学出版社) 第43—44页

    Levitan I B, Kaczmarek L K (translated by Shu S Y, Bao X M) 2001 The Neuron: Cell and Molecular Biology (Beijing: Science Press) pp43–44

    [32]

    Jokar E, Abolfathi H, Ahmadi A, Ahmadi M 2019 IEEE Trans. Circuits Syst. I Regul. Pap. 66 2336Google Scholar

    [33]

    Li K X, Bao H, Li H Z, Ma J, Hua Z Y, Bao B C 2021 IEEE Trans. Industr. Inform. 18 1726Google Scholar

    [34]

    Lin H R, Wang C H, Chen C J, Sun Y C, Zhou C, Xu C, Hong Q H 2021 IEEE Trans. Circuits Syst. I Regul. Pap. 68 3397Google Scholar

  • 图 1  G(x)与k1, m1, m2之间的关系图 (a) m1 = 3, m2 = –3时, k1G(x)之间的关系图; (b) k1 = 1, m2 = –3时, m1G(x)之间的关系图; (c) k1 = 1, m1 = 3时, m2G(x)之间的关系图

    Fig. 1.  Relationship between G(x) and k1, m1 and m2: (a) Relationship between k1 and G(x) when m1 = 3 and m2 = –3; (b) the relationship between m1 and G(x) when k1 = 1 and m2 = –3; (c) the relationship between m2 and G(x) when k1 = 1 and m1 = 3.

    图 2  忆阻器的滞回曲线和时域波形 (a) 不同幅值下的迟滞回线; (b) 不同频率下的迟滞回线; (c) 电压和电流的时域波形

    Fig. 2.  Hysteresis curves and time domain waveforms of memristor: (a) Hysteresis curves at different amplitudes; (b) hysteresis curves at different frequencies; (c) time domain waveforms of voltage and current.

    图 3  忆阻器的POP (a)和DC V-I图(b)

    Fig. 3.  POP (a) and DC V-I (b) diagram of the memristor.

    图 4  忆阻耦合的HNN的拓扑结构图

    Fig. 4.  Topological structure diagram of memristor coupled HNN.

    图 5  不同初始值下系统关于m1的分岔图 (a) (–2, 1, 1, 1, 6); (b) (–2, 1, 1, 1, 6.5); (c) (–2, 1, 1, 1, 8)

    Fig. 5.  Bifurcation diagram of the system about m1 under different initial values: (a) (–2, 1, 1, 1, 6); (b) (–2, 1, 1, 1, 6.5); (c) (–2, 1, 1, 1, 8).

    图 6  x1-x3平面共存非对称吸引子相图 (a) m1 = 0.01; (b) m1 = 0.05; (c) m1 = 0.1; (d) m1 = 0.83; (e) m1 = 1.13; (f) m1 = 2.5

    Fig. 6.  Coexistence asymmetric attractor phase diagram of the x1-x3 plane: (a) m1 = 0.01; (b) m1 = 0.05; (c) m1 = 0.1, (d) m1 = 0.83; (e) m1 = 1.13; (f) m1 = 2.5.

    图 7  系统的共存分岔图以及x1-x3平面共存对称吸引子相图 (a) 系统的共存分岔图; (b) m1 = 0.01; (c) m1 = 0.1

    Fig. 7.  Coexistence bifurcation diagram of the system and the coexistence symmetric attractor phase diagram of the x1-x3 plane: (a) Coexistence bifurcation diagram of the system; (b) m1 = 0.01; (c) m1 = 0.1.

    图 8  系统关于m2的分岔图(a)及李雅普诺夫指数谱(b)

    Fig. 8.  Bifurcation diagram (a) and Lyapunov exponent spectrum (b) of the system about m2.

    图 9  系统关于f的分岔图

    Fig. 9.  Bifurcation diagram of the system about f.

    图 10  不同频率下系统的相轨图(a), (c)和时域波形图(b), (d) (a), (b) f = 0.003; (c), (d) f = 0.1

    Fig. 10.  Phase diagram (a), (c) and time domain waveform diagram (b), (d) of the system at different frequencies: (a), (b) f = 0.003; (c), (d) f = 0.1.

    图 11  系统的实物连接图

    Fig. 11.  Physical connection diagram of the system.

    图 12  用STM32实现ODE45算法的流程图

    Fig. 12.  Flow chart of ODE45 algorithm implemented by STM32.

    图 13  示波器显示的相轨图(a)—(c)及对应Matlab仿真图(d)—(f) (a), (d) m1 = 0.02; (b), (e) m1 = 1.2; (c), (f) m1 = 2.5

    Fig. 13.  Oscilloscope display phase track diagram (a)–(c) and the corresponding Matlab simulation diagram (d)–(f): (a), (d) m1 = 0.02; (b), (e) m1 = 1.2; (c), (f) m1 = 2.5.

    表 1  a = 2.5, m1 = 0.5, m2 = –3时, 系统平衡点处的特征值及稳定性

    Table 1.  Eigenvalue and stability at the equilibrium point of the system when a = 2.5, m1 = 0.5 and m2 = –3.

    m1 k 特征值 稳定性
    0.5 0 1, 1.0399, –3.7032, 0.7316±3.9891i 不稳定的鞍焦点
    ±1 –1, 1.4681±2.5644i, –2.0681±40.1411i 不稳定的鞍焦点
    ±2 1, 1.4833±2.5761i, –2.0833±927.5014i 不稳定的鞍焦点
    ±3 –1, 1.4833±2.5761i, –2.0833±21462.9637i 不稳定的鞍焦点
    下载: 导出CSV

    表 2  不同的忆阻耦合Hopfield神经网络

    Table 2.  Different memristor-coupled Hopfield neural networks.

    系统 维度 局部有源
    忆阻器
    G(x)的类型 共存
    现象
    宽参数范围
    混沌状态
    文献[12] 三维 一次函数
    文献[14] 四维 双曲正切函数
    文献[15] 四维 二次函数
    文献[16] 二维 余弦函数
    文献[17] 四维 绝对值函数
    本文 五维 复合指数函数
    下载: 导出CSV
  • [1]

    Chua L 1971 IEEE Trans. Circuits Syst. I Regul. Pap. 18 507Google Scholar

    [2]

    Chua L 2014 Semicond. Sci. Technol. 29 104001Google Scholar

    [3]

    Wang M J, Deng Y, Liao X H, Li Z J, Ma M L, Zeng Y C 2019 Int. J. Non. Linear. Mech. 111 149Google Scholar

    [4]

    Peng Y X, Liu J, He S B, Sun K H 2023 Chaos, Solitons Fractals 171 113429Google Scholar

    [5]

    Wang M J, An M Y, He S B, Zhang X N, Iu H H, Li Z J 2023 Chaos 33 073129Google Scholar

    [6]

    Peng Y X, He S B, Sun K H 2022 Nonlinear Dyn. 107 1263Google Scholar

    [7]

    He S B, Liu J, Wang H H, Sun K H 2023 Neurocomputing (Amst. ) 523 1Google Scholar

    [8]

    古亚娜, 梁燕, 王光义, 夏晨阳 2022 物理学报 71 110501Google Scholar

    Gu Y N, Liang Y, Wang G Y, Xia C Y 2022 Acta Phys. Sin. 71 110501Google Scholar

    [9]

    武长春, 周莆钧, 王俊杰, 李国, 胡绍刚, 于奇, 刘洋 2022 物理学报 71 148401Google Scholar

    Wu C C, Zhou P J, Wang J J, Li G, Hu S G, Yu Q, Liu Y 2022 Acta Phys. Sin. 71 148401Google Scholar

    [10]

    Chua L 2013 Nanotechnology 24 383001Google Scholar

    [11]

    Lu Y C, Li H M, Li C L 2023 Neurocomputing (Amst. ) 544 126246Google Scholar

    [12]

    Li Z J, Chen K J 2023 Chaos, Solitons Fractals 175 114017Google Scholar

    [13]

    Hopfield J J 1984 Proc. Natl. Acad. Sci. U. S. A. 81 3088Google Scholar

    [14]

    Chen C J, Chen J Q, Bao H, Chen M, Bao B C 2019 Nonlinear Dyn. 95 3385Google Scholar

    [15]

    Lin H R, Wang C H, Hong Q H, Sun Y C 2020 IEEE Tran. Circuits Syst. II Express Briefs 67 3472Google Scholar

    [16]

    Chen C J, Min F H, Zhang Y Z, Bao B C 2021 Nonlinear Dyn. 106 2559Google Scholar

    [17]

    Li C L, Yang Y Y, Yang X B, Zi X Y, Xiao F L 2022 Nonlinear Dyn. 108 1697Google Scholar

    [18]

    Doubla I S, Ramakrishnan B, Njitacke Z T, Kengne J, Rajagopal K 2022 Int. J. Electron. Commun. 144 154059Google Scholar

    [19]

    黄丽丽, 黄强, 黄振, 臧红岩, 雷腾飞 2023 电子元件与材料 42 10Google Scholar

    Huang L L, Huang Q, Huang Z, Zang H Y, Lei T F 2023 Electron. Compon. Mater. 42 10Google Scholar

    [20]

    Lin H R, Wang C H, Sun J R, Zhang X, Sun Y C, Iu H H C 2023 Chaos Solitons Fractals 166 112905Google Scholar

    [21]

    Wan Q Z, Chen S M, Yang Q, Liu J, Sun K L 2023 Nonlinear Dyn. 111 18505Google Scholar

    [22]

    Lin H R, Wang C H, Yu F, Sun J R, Du S C, Deng Z K, Deng Q L 2023 Mathematics (Basel) 11 1369Google Scholar

    [23]

    Panahi S, Aram Z, Jafari S, Ma J, Sprott J C 2017 Chaos, Solitons Fractals 105 150Google Scholar

    [24]

    Guevara M R, Glass L, Mackey M C, Shrier A 1983 IEEE Trans. Syst. Man Cybern. Syst. 5 790Google Scholar

    [25]

    Chua L 2018 Appl. Phys. A: Mater. 124 563Google Scholar

    [26]

    Chua L 2005 Int. J. Bifurcat. Chaos 15 3435Google Scholar

    [27]

    Ascoli A, Slesazeck S, Mähne H, Tetzlaff R, Mikolajick T 2015 IEEE Trans. Circuits Syst. I Regul. Pap. 62 1165Google Scholar

    [28]

    Chua L 2015 Radioengineering 24 319Google Scholar

    [29]

    Wang M J, Li J H, Yu S S, Zhang X N, Li Z J, Iu H H C 2020 Chaos 30 043125Google Scholar

    [30]

    Bi Q S, Gou J T 2023 Chaos Solitons Fractals 167 113046Google Scholar

    [31]

    莱维坦 I B, 卡茨玛克 L K著 (舒斯云, 包新民 译) 2001 神经元: 细胞和分子生物学 (北京: 科学出版社) 第43—44页

    Levitan I B, Kaczmarek L K (translated by Shu S Y, Bao X M) 2001 The Neuron: Cell and Molecular Biology (Beijing: Science Press) pp43–44

    [32]

    Jokar E, Abolfathi H, Ahmadi A, Ahmadi M 2019 IEEE Trans. Circuits Syst. I Regul. Pap. 66 2336Google Scholar

    [33]

    Li K X, Bao H, Li H Z, Ma J, Hua Z Y, Bao B C 2021 IEEE Trans. Industr. Inform. 18 1726Google Scholar

    [34]

    Lin H R, Wang C H, Chen C J, Sun Y C, Zhou C, Xu C, Hong Q H 2021 IEEE Trans. Circuits Syst. I Regul. Pap. 68 3397Google Scholar

  • [1] 王璇, 杜健嵘, 李志军, 马铭磷, 李春来. 串扰忆阻突触异质离散神经网络的共存放电与同步行为. 物理学报, 2024, 73(11): 110503. doi: 10.7498/aps.73.20231972
    [2] 贾美美, 曹佳伟, 白明明. 新型忆阻耦合异质神经元的放电模式和预定义时间混沌同步. 物理学报, 2024, 73(17): 170502. doi: 10.7498/aps.73.20240872
    [3] 全旭, 邱达, 孙智鹏, 张贵重, 刘嵩. 一个具有共存吸引子的四阶混沌系统动力学分析及FPGA实现. 物理学报, 2023, 72(19): 190502. doi: 10.7498/aps.72.20230795
    [4] 古亚娜, 梁燕, 王光义, 夏晨阳. NbOx忆阻神经元的设计及其在尖峰神经网络中的应用. 物理学报, 2022, 71(11): 110501. doi: 10.7498/aps.71.20220141
    [5] 武长春, 周莆钧, 王俊杰, 李国, 胡绍刚, 于奇, 刘洋. 基于忆阻器的脉冲神经网络硬件加速器架构设计. 物理学报, 2022, 71(14): 148401. doi: 10.7498/aps.71.20220098
    [6] 秦铭宏, 赖强, 吴永红. 具有无穷共存吸引子的简单忆阻混沌系统的分析与实现. 物理学报, 2022, 71(16): 160502. doi: 10.7498/aps.71.20220593
    [7] 王世场, 卢振洲, 梁燕, 王光义. N型局部有源忆阻器的神经形态行为. 物理学报, 2022, 71(5): 050502. doi: 10.7498/aps.71.20212017
    [8] 张绍华, 王聪, 张宏立. 永磁同步电动机的簇发振荡分析及协同控制. 物理学报, 2020, 69(21): 210501. doi: 10.7498/aps.69.20200413
    [9] 张正娣, 刘亚楠, 李静, 毕勤胜. 分段Filippov系统的簇发振荡及擦边运动机理. 物理学报, 2018, 67(11): 110501. doi: 10.7498/aps.67.20172421
    [10] 张正娣, 刘杨, 张苏珍, 毕勤胜. 余维-1非光滑分岔下的簇发振荡及其机理. 物理学报, 2017, 66(2): 020501. doi: 10.7498/aps.66.020501
    [11] 吴天一, 陈小可, 张正娣, 张晓芳, 毕勤胜. 非对称型簇发振荡吸引子结构及其机理分析. 物理学报, 2017, 66(11): 110501. doi: 10.7498/aps.66.110501
    [12] 包涵, 包伯成, 林毅, 王将, 武花干. 忆阻自激振荡系统的隐藏吸引子及其动力学特性. 物理学报, 2016, 65(18): 180501. doi: 10.7498/aps.65.180501
    [13] 邢雅清, 陈小可, 张正娣, 毕勤胜. 多平衡态下簇发振荡产生机理及吸引子结构分析. 物理学报, 2016, 65(9): 090501. doi: 10.7498/aps.65.090501
    [14] 刘玉东, 王连明. 基于忆阻器的spiking神经网络在图像边缘提取中的应用. 物理学报, 2014, 63(8): 080503. doi: 10.7498/aps.63.080503
    [15] 李志军, 曾以成, 李志斌. 改进型细胞神经网络实现的忆阻器混沌电路. 物理学报, 2014, 63(1): 010502. doi: 10.7498/aps.63.010502
    [16] 张昀, 张志涌, 于舒娟. 基于幅值相位型离散Hopfield神经网络的多进制振幅键控盲检测. 物理学报, 2012, 61(14): 140701. doi: 10.7498/aps.61.140701
    [17] 张昀, 张志涌. 复数多值离散Hopfield神经网络的稳定性研究. 物理学报, 2011, 60(9): 090703. doi: 10.7498/aps.60.090703
    [18] 包伯成, 刘中, 许建平, 朱雷. 基于Colpitts振荡器模型生成的多涡卷超混沌吸引子. 物理学报, 2010, 59(3): 1540-1548. doi: 10.7498/aps.59.1540
    [19] 熊 涛, 张便利, 常胜江, 申金媛, 张延忻. 基于Hopfield神经网络的信元调度多重输入队列ATM交换结构及算法. 物理学报, 2005, 54(5): 2435-2440. doi: 10.7498/aps.54.2435
    [20] 马余强, 张玥明, 龚昌德. Hopfield神经网络模型的恢复特性. 物理学报, 1993, 42(8): 1356-1360. doi: 10.7498/aps.42.1356
计量
  • 文章访问数:  1740
  • PDF下载量:  70
  • 被引次数: 0
出版历程
  • 收稿日期:  2023-11-30
  • 修回日期:  2024-04-29
  • 上网日期:  2024-05-17
  • 刊出日期:  2024-07-05

/

返回文章
返回