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

基于忆容器件的神经形态计算研究进展

任宽 张珂嘉 秦溪子 任焕鑫 朱守辉 杨峰 孙柏 赵勇 张勇

引用本文:
Citation:

基于忆容器件的神经形态计算研究进展

任宽, 张珂嘉, 秦溪子, 任焕鑫, 朱守辉, 杨峰, 孙柏, 赵勇, 张勇

Research progress of neuromorphic computation based on memcapacitors

Ren Kuan, Zhang Ke-Jia, Qin Xi-Zi, Ren Huan-Xin, Zhu Shou-Hui, Yang Feng, Sun Bai, Zhao Yong, Zhang Yong
PDF
HTML
导出引用
  • 人工智能的快速发展需要人工智能专用硬件的快速发展, 受人脑存算一体、并行处理启发而构建的包含突触与神经元的神经形态计算架构, 可以有效地降低人工智能中计算工作的能耗. 记忆元件在神经形态计算的硬件实现中展现出巨大的应用价值; 相比传统器件, 用忆阻器构建突触、神经元能极大地降低计算能耗, 然而在基于忆阻器构建的神经网络中, 更新、读取等操作存在由忆阻电压电流造成的系统性能量损失. 忆容器作为忆阻器衍生器件, 被认为是实现低耗能神经网络的潜在器件, 引起国内外研究者关注. 本文综述了实物/仿真忆容器件及其在神经形态计算中的最新进展, 主要包括目: 前实物/仿真忆容器原理与特性, 代表性的忆容突触、神经元及神经形态计算架构, 并通过总结近年来忆容器研究所取得的成果, 对当前该领域面临的挑战及未来忆容神经网络发展的重点进行总结与展望.
    The rapid development of artificial intelligence (AI) requires one to speed up the development of the domain-specific hardware specifically designed for AI applications. The neuromorphic computing architecture consisting of synapses and neurons, which is inspired by the integrated storage and parallel processing of human brain, can effectively reduce the energy consumption of artificial intelligence in computing work. Memory components have shown great application value in the hardware implementation of neuromorphic computing. Compared with traditional devices, the memristors used to construct synapses and neurons can greatly reduce computing energy consumption. However, in neural networks based on memristors, updating and reading operations have system energy loss caused by voltage and current of memristors. As a derivative of memristor, memcapacitor is considered as a potential device to realize a low energy consumption neural network, which has attracted wide attention from academia and industry. Here, we review the latest advances in physical/simulated memcapacitors and their applications in neuromorphic computation, including the current principle and characteristics of physical/simulated memcapacitor, representative synapses, neurons and neuromorphic computing architecture based on memcapacitors. We also provide a forward-looking perspective on the opportunities and challenges of neuromorphic computation based on memcapacitors.
      通信作者: 杨峰, yf@swjtu.edu.cn ; 张勇, yongzhang@swjtu.edu.cn
    • 基金项目: 国家高技术研究发展计划(批准号: 2017YFE0301401)资助的课题
      Corresponding author: Yang Feng, yf@swjtu.edu.cn ; Zhang Yong, yongzhang@swjtu.edu.cn
    • Funds: Project supported by the National High Technology Research and Development Program of China (Grant No. 2017YFE0301401)
    [1]

    Goodfellow I, Bengio Y, Courville A 2016 Deep Learning (Cambridge: The MIT Press) pp1−100

    [2]

    James C D, Aimone J B, Miner N E, Vineyard C M, Rothganger F H, Carlson K D, Mulder S A, Draelos T J, Faust A, Marinella M J, Naegle J H, Plimpton S J 2017 Biol. Inspired Cogn. Archit. 19 49Google Scholar

    [3]

    Merolla P A, Arthur J V, Alvarez-Icaza R, Cassidy A S, Sawada J, Akopyan F, Jackson B L, Imam N, Guo C, Nakamura Y, Brezzo B, Vo I, Esser S K, Appuswamy R, Taba B, Amir A, Flickner M D, Risk W P, Manohar R, Modha D S 2014 Science 345 668Google Scholar

    [4]

    Furber S B, Galluppi F, Temple S, Plana L A 2014 Proc. IEEE 102 652Google Scholar

    [5]

    Chua L 1971 IEEE Trans. Circuit Theory 18 507Google Scholar

    [6]

    Strukov D B, Snider G S, Stewart D R, Williams R S 2008 Nature 453 80Google Scholar

    [7]

    Dev D, Krishnaprasad A, Shawkat M S, He Z, Das S, Fan D, Chung H S, Jung Y, Roy T 2020 IEEE Electron Device Lett. 41 936Google Scholar

    [8]

    He C, Tang J, Shang D S, Tang J, Xi Y, Wang S, Li N, Zhang Q, Lu J K, Wei Z, Wang Q, Shen C, Li J, Shen S, Shen J, Yang R, Shi D, Wu H, Wang S, Zhang G 2020 ACS Appl. Mater. Interfaces 12 11945Google Scholar

    [9]

    Wang H, Yan X B, Zhao M L, Zhao J H, Zhou Z Y, Wang J J, Hao W C 2020 Appl. Phys. Lett. 116 093501Google Scholar

    [10]

    Chen J R, Wu H Q, Gao B, Tang J S, Hu X B S, Qian H 2020 IEEE Trans. Electron Devices 67 2213Google Scholar

    [11]

    Liao Y, Gao B, Xu F, Yao P, Chen J R, Zhan W Q, Tang J S, Wu H Q, Qian H 2020 IEEE Trans. Electron Devices 67 1593Google Scholar

    [12]

    Yao P, Wu H, Gao B, Tang J, Zhang Q, Zhang W, Yang J J, Qian H 2020 Nature 577 641Google Scholar

    [13]

    Li X, Tang J, Zhang Q, Gao B, Yang J J, Song S, Wu W, Zhang W, Yao P, Deng N, Deng L, Xie Y, Qian H, Wu H 2020 Nat. Nanotechnol. 15 776Google Scholar

    [14]

    Di Ventra M, Pershin Y V, Chua L O 2009 Proc. IEEE 97 1717Google Scholar

    [15]

    Flak J 2012 13th International Workshop on Cellular Nanoscale Networks and their Applications Turin, Italy, Aug. 29−31 2012 p1

    [16]

    Fouda M E, Radwan A G 2014 26th International Conference on Microelectronics (ICM) Doha, Qatar, Dec. 14−17 2014 p172

    [17]

    Pershin Y V, Di Ventra M 2014 Electron. Lett. 50 141Google Scholar

    [18]

    Yi S, ZhenZhen J, XiaoPing W, Yang L 2015 34th Chinese Control Conference (CCC) Hangzhou, China, July 28–30 2015 p3452

    [19]

    Tran S J D, Teuscher C 2017 IEEE/ACM International Symposium on Nanoscale Architectures (NANOARCH) Newport, RI, July 25−26 2017 p115

    [20]

    Wang Z, Rao M, Han J W, Zhang J, Lin P, Li Y, Li C, Song W, Asapu S, Midya R, Zhuo Y, Jiang H, Yoon J H, Upadhyay N K, Joshi S, Hu M, Strachan J P, Barnell M, Wu Q, Wu H, Qiu Q, Williams R S, Xia Q, Yang J J 2018 Nat Commun. 9 3208Google Scholar

    [21]

    Chen Y, Zhang J, Zhang Y, Zhang R, Kimura M, Nakashima Y 2019 17th IEEE International New Circuits and Systems Conference (NEWCAS) Munich, Germany, June 23−26 2019 p1

    [22]

    Tran S J D, Teuscher C 2019 IEEE International Conference on Rebooting Computing (ICRC) San Mateo, CA, Nov. 6−8 2019 p110

    [23]

    L.Chua 2015 Radioengineering 24 319Google Scholar

    [24]

    Bessonov A A, Kirikova M N, Petukhov D I, Allen M, Ryhanen T, Bailey M J 2015 Nat. Mater. 14 199Google Scholar

    [25]

    Goswami S, Rath S P, Thompson D, Hedstrom S, Annamalai M, Pramanick R, Ilic B R, Sarkar S, Hooda S, Nijhuis C A, Martin J, Williams R S, Goswami S, Venkatesan T 2020 Nat. Nanotechnol. 15 380Google Scholar

    [26]

    Lai Q X, Zhang L, Li Z Y, Stickle W F, Williams R S, Chen Y 2009 Appl. Phys. Lett. 95 213503Google Scholar

    [27]

    Liu R X, Dong R X, Qin S C, Yan X L 2020 Org. Electron. 81 105680Google Scholar

    [28]

    Liu S Q, Wu N J, Ignatiev A, Li J R 2006 J. Appl. Phys. 100 056101Google Scholar

    [29]

    Martinez-Rincon J, Di Ventra M, Pershin Y V 2010 Phys. Rev. B. 81 195430Google Scholar

    [30]

    Najem J S, Hasan M S, Williams R S, Weiss R J, Rose G S, Taylor G J, Sarles S A, Collier C P 2019 Nat Commun. 10 3239Google Scholar

    [31]

    Nieminen H, Ermolov V, Nybergh K, Silanto S, Ryhanen T 2002 J. Micromech. Microeng. 12 177Google Scholar

    [32]

    Noh Y J, Baek Y J, Hu Q, Kang C J, Choi Y J, Lee H H, Yoon T S 2015 IEEE Trans. Nanotechnol. 14 798Google Scholar

    [33]

    Park D, Yang P, Kim H J, Beom K, Lee H H, Kang C J, Yoon T S 2018 Appl. Phys. Lett. 113 162102Google Scholar

    [34]

    Román Acevedo W, van den Bosch C A M, Aguirre M H, Acha C, Cavallaro A, Ferreyra C, Sánchez M J, Patrone L, Aguadero A, Rubi D 2020 Appl. Phys. Lett. 116 063502Google Scholar

    [35]

    Salaoru I, Khiat A, Li Q J, Berdan R, Prodromakis T 2013 Appl. Phys. Lett. 103 233513Google Scholar

    [36]

    Slesazeck S, Wylezich H, Mikolajick T 2017 IEEE 8th Latin American Symposium on Circuits & Systems (LASCAS) Bariloche, Argentina, Feb. 20−23 2017 p1

    [37]

    Sun J, Lind E, Maximov I, Xu H Q 2011 IEEE Electron Device Lett. 32 131Google Scholar

    [38]

    Wu S X, Peng H Y, Wu T 2011 Appl. Phys. Lett. 98 093503Google Scholar

    [39]

    Ahmed M G, Cho K, Cho T 2012 13th International Workshop on Cellular Nanoscale Networks and their Applications Turin, Italy, Aug. 29−31 2012 p1

    [40]

    Asapu S, Pershin Y V 2015 IEEE Trans. Electron Devices 62 3678Google Scholar

    [41]

    Biolek D, Biolek Z, Biolkova V 2009 European Conference on Circuit Theory and Design Antalya, Turkey, Aug. 23−27 2009 p249

    [42]

    Biolek D, Biolek Z, Biolkova V 2010 Electron. Lett. 46 520Google Scholar

    [43]

    Biolek D, Biolkova V 2010 Electron. Lett. 46 1428Google Scholar

    [44]

    Biolek D, Biolková V, Kolka Z 2010 IEEE Asia Pacific Conference on Circuits and Systems Kuala Lumpur, Malaysia Dec. 6−9 2010 p800

    [45]

    Flak J, Raantala A, Haatainen T, Prunnila M, Laiho M 2014 14th International Workshop on Cellular Nanoscale Networks and their Applications (CNNA) Notre Dame, IN, USA, July 29−31 2014 p1

    [46]

    Fouda M E, Radwan A G 2012 Electron. Lett. 48 1454Google Scholar

    [47]

    Pershin Y V, Di Ventra M 2010 Electron. Lett. 46 517Google Scholar

    [48]

    Pershin Y V, Di Ventra M 2011 Electron. Lett. 47 243Google Scholar

    [49]

    Romero F J, Morales D P, Godoy A, Ruiz F G, Tienda-Luna I M, Ohata A, Rodriguez N 2019 Int. J. Circ. Theor. App. 47 572Google Scholar

    [50]

    Yu D S, Liang Y, Iu H H C, Chua L O 2014 IEEE Trans. Circuits Syst. II-Express Briefs 61 758Google Scholar

    [51]

    Yu D, Zhao X, Sun T, Iu H H C, Fernando T 2020 IEEE Trans. Circuits Syst. II-Express Briefs 67 1334Google Scholar

    [52]

    Yu D, Zhou Z, Iu H H C, Fernando T, Hu Y 2016 IEEE Trans. Circuits Syst. II-Express Briefs 63 1101Google Scholar

    [53]

    Yu D S, Liang Y, Chen H, Iu H H C 2013 IEEE Trans. Circuits Syst. II-Express Briefs 60 207Google Scholar

    [54]

    Zheng C Y, Yu D S, Iu H H C, Fernando T, Sun T T, Eshraghian J K, Guo H D 2019 IEEE Trans. Circuits Syst. I-Regul. Pap. 66 4793Google Scholar

    [55]

    Kwon D, Chung I Y 2020 IEEE Electron Device Lett. 41 493Google Scholar

    [56]

    Zhao L, Fan Z, Cheng S L, Hong L Q, Li Y Q, Tian G, Chen D Y, Hou Z P, Qin M H, Zeng M, Lu X B, Zhou G F, Gao X S, Liu J M 2020 Adv Electron Mater 6 1900858Google Scholar

    [57]

    Yamaletdinov R D, Ivakhnenko O V, Sedelnikova O V, Shevchenko S N, Pershin Y V 2018 Sci. Rep. 8 3566Google Scholar

    [58]

    Patel J A, Sandhie Z T, Chowdhury M H 2018 IEEE 61st International Midwest Symposium on Circuits and Systems (MWSCAS) Windsor, Canada, Aug. 5–8 2018 p1130

    [59]

    Salaoru I, Li Q, Khiat A, Prodromakis T 2014 Nanoscale. Res. Lett. 9 552Google Scholar

    [60]

    Cai J W, Li L X, Xu C, Feng Y, Zhong Y N, Xu J L, Gao X, Wang S D 2019 Appl. Phys. Lett. 114 043302Google Scholar

    [61]

    Qian W H, Cheng X F, Zhao Y Y, Zhou J, He J H, Li H, Xu Q F, Li N J, Chen D Y, Lu J M 2019 Adv. Mater. 31 1806424Google Scholar

    [62]

    Yang P, Jun Kim H, Zheng H, Won Beom G, Park J S, Jung Kang C, Yoon T S 2017 Nanotechnology 28 225201Google Scholar

    [63]

    Martinez-Rincon J, Pershin Y V 2011 IEEE Trans. Electron Devices 58 1809Google Scholar

    [64]

    Yang C, Yang N, Yu Y, Li Y, Diez F F 2017 IEEE 17th International Conference on Communication Technology (ICCT) Chengdu, China, Oct. 27–30 2017 p1171

    [65]

    Corinto F, Di Marco M, Forti M, Chua L 2019 IEEE Trans Cybern 50 4758Google Scholar

    [66]

    Cohen G Z, Pershin Y V, Di Ventra M 2012 Phys. Rev. B. 85 165428Google Scholar

    [67]

    Mcculloch W S, Pitts W 1943 Bull. Math. Biol. 5 115

    [68]

    Hodgkin A L, Huxley A F 1989 Bull. Math. Biol. 52 25

    [69]

    Pershin Y V, Di Ventra M 2011 Adv. Phys. 60 145Google Scholar

    [70]

    Rumelhart D E, Hinton G E, Williams R J 1986 Nature 323 533Google Scholar

    [71]

    John H, Anders K, Palmer R G 1991 Phys. Today 44 70

    [72]

    Bi G Q, Poo M M 2001 Annu. Rev. Neurosci. 24 139Google Scholar

  • 图 1  (a)忆容系统的捏滞曲线[14]; (b)仿真的压控忆容器q-V曲线[14]; (c) 仿真的压控忆容器C-V曲线[14]

    Fig. 1.  (a) Schematics of a pinched hysteresis loop of a memcapacitive system[14]; (b) q-V curve of a simulated voltage-controlled memcapacitor[14]; (c) C-V curve of a simulated voltage-controlled memcapacitor[14].

    图 2  基于忆容的神经形态计算

    Fig. 2.  Neuromorphic computation based on memcapacitors.

    图 3  ITO (In-Sn-O)/HfOx/p-Si结构忆容器及C-V曲线[33]

    Fig. 3.  Structure of ITO (In-Sn-O)/HfOx/p-S memcapacitor and its C-V curves[33].

    图 4  Au/Ti/HfOx/InP结构忆容器[37] (a) 器件结构及总I-V曲线; (b) 零偏压下器件能带结构; (c) 器件$ {\rm{R}}{\rm{C}} $等效电路及C-V曲线; (d) 器件R-V曲线

    Fig. 4.  Structure of Au/Ti/HfOx/InP memcapacitor[37]: (a) device structure and total I-V curves; (b) schematics for the band diagram of the metal HfO2-semiconductordiode at zero bias; (c) equivalent circuit of device and its C-V curves; (d) R-V curves.

    图 5  室温下Pt/Pr0.7Ca0.3MnO3(PCMO)/YBCO/LAO结构忆容器性质[28] (a) 非易失电容随脉冲电压数的变化; (b) 非易失电阻随脉冲电压数的变化; (c) 非易失电容随测试电压频率的变化

    Fig. 5.  Nonvolatile capacitance and resistance changes for Au/PCMO/YBCO/LAO structure sample at room temperature[28]: (a) Nonvolatile capacitance changes with applied pulse numbers; (b) nonvolatile resistance changes with applied pulse numbers; (c) nonvolatile capacitance changes with frequency.

    图 6  (a) Ag(TE)/MoOx/MoS2/Ag(BE)忆容器件结构[24]; (b)钼氧化态MoOx/MoS2样品在200 ℃持续3 h退火后的XPS剖面; 填充区域代表一个Mo6 +丰富的区域[24]; (c)电阻、电容开关性质[24]

    Fig. 6.  [24](a) Ag(TE)/MoOx/MoS2/Ag(BE) structure memcapacitor[24]; (b) Molybdenum oxidation-state XPS profile of the MoOx/MoS2 sample annealed at 200 ℃ for 3 h; the filled area represents a Mo6+ -rich region[24]; (c) capacitance and resistance switch characteristics[24].

    图 7  (a)Al/Ti/RbAg4I5 /MEH-PPV/SiO2 /p-Si/Al忆容器结构及其特性曲线[26]; (b) ITO/MASnBr3/Au结构图[61]; (c) ITO/MASnBr3/Au原理图[61]; (d) ITO/MASnBr3/Au结构的忆容特性(1 MHz下的C-VQ-V特性)[61]; (e) 硅底电极保持接地的有机薄膜记忆电容的器件结构和正负偏压下的电荷积累方案[60]

    Fig. 7.  (a) A memory capacitor with an Al/Ti/RbAg4I5 /MEH-PPV/SiO2 /p-Si/Al structure (inset) and its characteristic curve[26]; (b) schematic diagram of the ITO/MASnBr3/Au structure[61]; (c) mechanism of the ITO/MASnBr3/Au structure[61]; (d) memcapacitive characteristics of the ITO/MASnBr3/Au device(C-V hysteresis and Q-V loops detected at 1 MHz); (e) device structure and charge accumulation scheme under negative (top) and positive (bottom) biases of an organic thin film memcapacitor, where the Si bottom electrode is kept grounded[60].

    图 8  基于记忆电容的人工突触短期塑性模拟[27] (a) 生物突触和Al/共聚物/ITO人工突触装置信号传输示意图, 共聚物薄膜的AFM图像; (b) C-V曲线; (c) 器件的PPF行为, A1和A2分别代表第一个和第二个突触前突起的PSC, 红色和蓝色曲线分别代表正、负电压下的兴奋性PSC和抑制性PSC; (d) PPF指数被绘制成时间间隔的函数

    Fig. 8.  Short-term plasticity emulated in artificial synapse based on memory capacitance[27]: (a) Schematic illustrations of the signal transmission in biological synapse and the Al/copolymer/ITO artificial synaptic device. AFM image of copolymer film; (b) the C-V curves; (c) PPF behaviors of the device. A1 and A2 represent the PSC of the first and second presynaptic spike, respectively. The red and blue curves represent the excitatory and inhibitory PSC under negative and positive voltage, respectively. The inset shows schematic of pulse application; (d)PPF index plotted as a function of the time interval.

    图 9  带[Ru(L)2](PF6)2层器件的测试结构及电学特性[25] (a) 3种结构的示意图; (b)—(d) A(b), B (c)和C (d)结构电流密度对电压J(V)的特性; (e)—(f) 顶部面板显示了结构A(e)和B(f)的相对介电常数与电压的特性, 并覆盖了相应结构的J(V)曲线; 底部的面板显示了结构A(e)和B(f)对应的电荷和电压曲线

    Fig. 9.  Test structures and electrical characterizations of devices with [Ru(L)2](PF6)[25]: (a) Schematic illustration of the three structures; (b)–(d) the current density versus voltage J(V) characteristics of structures A(b), B(c) and C(d); (e)–(f) the top panels show the relative permittivity versus voltage characteristics of structures A(e) and B(f), overlaid with the J(V) curves of the corresponding structures. The bottom panels show the corresponding charge versus voltage profiles for structures A(e) and B(f).

    图 10  电子隧穿忆容模型[29]

    Fig. 10.  Scheme of an electron tunneling memcapacitor[29].

    图 11  (a) 双状态MEM忆容[31]; (b) 忆容的电容-电压曲线[31]; (c) 弹性电极忆容

    Fig. 11.  (a) Photograph of a two-state MEM capacitor[31]; (b) measured capacitance as a function of voltage of the two-state capacitor[31]; (c) elastic poles memcapacitor.

    图 12  生物忆容器仿生膜组装与电行为[30] (a) 一种模拟生物膜结构的电容平面脂质双分子层, 在脂质包被的微滴之间接触并排除多余油脂后自发形成; (b) 由静膜电压v(t)引起的几何变化示意图

    Fig. 12.  Biomimetic membrane assembly and electromechanical behaviours[30]: (a) A capacitive planar lipid bilayer that mimics the structure of a biological membrane forms spontaneously upon contact between lipid-coated droplets and exclusion of excess oil; (b) a schematic describing the geometrical changes caused by a net membrane voltage, v(t).

    图 13  伪忆容[20] (a) 伪忆容的扫描电子显微图的平面视图和透射电子显微图的截面图; (b)集成伪忆容的电荷-电压关系

    Fig. 13.  Dynamic pseudo-memcapacitor(DPM)[20]: (a) a scanning electron micrograph of the plan view of the integrated DPM, and a transmission electron micrograph of the cross-section; (b) charge-voltage relationship of the integrated DPM.

    图 14  忆容仿真电路原理图 (a) 基于密勒效应的忆容仿真电路[49]; (b) 忆容-电阻串联电路[47]; (c) 提出的多功能电路[51]

    Fig. 14.  Schematic of the memcapacitor emulator: (a) Schematic of the memcapacitor emulator based on the Miller effect[49]; (b) memcapacitor-resistor series circuit[47]; (c) the proposed mutator circuit[51].

    图 15  MP神经元模型

    Fig. 15.  MP-neuron mode.

    图 16  忆容桥式突触电路[16]

    Fig. 16.  Memcapacitor bridge synaptic circuit[16].

    图 17  忆容记忆突触[17] (a) 集成神经网络中的忆容突触; (b) 忆容突触实现STDP; (c) 单个突触的集成忆容神经元点火仿真

    Fig. 17.  Memcapacitive synapses[17] (a) Memcapacitive synapses in integrate-and-fire neural network; (b) STDP with memcapacitive synapses; (c) simulation of integrate-and-fire memcapacitive network with only one spiking neuron.

    图 18  伪忆容突触[20] (a) 生物神经元接受高频突触后输入后产生动作电位的示意图; (b) 伪忆容的集成和触发过程; (c) 电子神经元晶体管的原理图; (d) 电子神经元-晶体管集成-点火过程的动力学

    Fig. 18.  Pseudo-memcapacitor synapse[20]: (a) Schematic representation of a biological neuron generating an action potential after receiving high-frequency post-synaptic inputs; (b) the integrate-and-fire process of a pseudo-memcapacitor synapse; (c) schematic of the synapse-transistor; (d) dynamics of the synapse-transistor integrate-and-fire process.

    图 19  忆容-MOS耦合神经元胞体[21]

    Fig. 19.  Neuron-MOS transistor couples the memcapacitor cells[21].

    图 20  MC-ACU[21] (a) 全结构电路图; (b) sigmoid神经元电路; (c) 在HSPICE中的仿真曲线(蓝)与理论数学曲线(红)对比; (d) 线性神经元电路; (e)在HSPICE中的仿真曲线(蓝)与理论数学曲线(红)对比

    Fig. 20.  MC-ACU[21]: (a) Overall architecture; (b) sigmoid neuron circuit; (c) simulation results in HSPICE (blue) compared with the mathematical sigmoid(red); (d) linear neuron circuit; (e) simulation results in HSPICE (blue) compared with the mathematical linear (red).

    图 21  基于电容式网络的联想学习机制[20]. 两个突触前信号分别模拟食物的视觉和铃声. 突触后神经元模拟狗的唾液分泌. 与“食物”突触前神经元连接的突触的初始权重较大, 而与“钟”突触前神经元连接的突触的初始权重较小

    Fig. 21.  Capacitive network for associative learning based on the Hebbian-like mechanism[20]. Two pre-synaptic signals model the sight of food and the sound of a bell, respectively. The post-synaptic neuron models the salivation of a dog. The initial weight of the synapse interfacing with the “food” pre-synaptic neuron was large, while that of the synapse connected to the “bell” pre-synaptic neuron was small

    图 22  忆容储层计算网络[22]

    Fig. 22.  A memcapacitive reservoir network[22].

  • [1]

    Goodfellow I, Bengio Y, Courville A 2016 Deep Learning (Cambridge: The MIT Press) pp1−100

    [2]

    James C D, Aimone J B, Miner N E, Vineyard C M, Rothganger F H, Carlson K D, Mulder S A, Draelos T J, Faust A, Marinella M J, Naegle J H, Plimpton S J 2017 Biol. Inspired Cogn. Archit. 19 49Google Scholar

    [3]

    Merolla P A, Arthur J V, Alvarez-Icaza R, Cassidy A S, Sawada J, Akopyan F, Jackson B L, Imam N, Guo C, Nakamura Y, Brezzo B, Vo I, Esser S K, Appuswamy R, Taba B, Amir A, Flickner M D, Risk W P, Manohar R, Modha D S 2014 Science 345 668Google Scholar

    [4]

    Furber S B, Galluppi F, Temple S, Plana L A 2014 Proc. IEEE 102 652Google Scholar

    [5]

    Chua L 1971 IEEE Trans. Circuit Theory 18 507Google Scholar

    [6]

    Strukov D B, Snider G S, Stewart D R, Williams R S 2008 Nature 453 80Google Scholar

    [7]

    Dev D, Krishnaprasad A, Shawkat M S, He Z, Das S, Fan D, Chung H S, Jung Y, Roy T 2020 IEEE Electron Device Lett. 41 936Google Scholar

    [8]

    He C, Tang J, Shang D S, Tang J, Xi Y, Wang S, Li N, Zhang Q, Lu J K, Wei Z, Wang Q, Shen C, Li J, Shen S, Shen J, Yang R, Shi D, Wu H, Wang S, Zhang G 2020 ACS Appl. Mater. Interfaces 12 11945Google Scholar

    [9]

    Wang H, Yan X B, Zhao M L, Zhao J H, Zhou Z Y, Wang J J, Hao W C 2020 Appl. Phys. Lett. 116 093501Google Scholar

    [10]

    Chen J R, Wu H Q, Gao B, Tang J S, Hu X B S, Qian H 2020 IEEE Trans. Electron Devices 67 2213Google Scholar

    [11]

    Liao Y, Gao B, Xu F, Yao P, Chen J R, Zhan W Q, Tang J S, Wu H Q, Qian H 2020 IEEE Trans. Electron Devices 67 1593Google Scholar

    [12]

    Yao P, Wu H, Gao B, Tang J, Zhang Q, Zhang W, Yang J J, Qian H 2020 Nature 577 641Google Scholar

    [13]

    Li X, Tang J, Zhang Q, Gao B, Yang J J, Song S, Wu W, Zhang W, Yao P, Deng N, Deng L, Xie Y, Qian H, Wu H 2020 Nat. Nanotechnol. 15 776Google Scholar

    [14]

    Di Ventra M, Pershin Y V, Chua L O 2009 Proc. IEEE 97 1717Google Scholar

    [15]

    Flak J 2012 13th International Workshop on Cellular Nanoscale Networks and their Applications Turin, Italy, Aug. 29−31 2012 p1

    [16]

    Fouda M E, Radwan A G 2014 26th International Conference on Microelectronics (ICM) Doha, Qatar, Dec. 14−17 2014 p172

    [17]

    Pershin Y V, Di Ventra M 2014 Electron. Lett. 50 141Google Scholar

    [18]

    Yi S, ZhenZhen J, XiaoPing W, Yang L 2015 34th Chinese Control Conference (CCC) Hangzhou, China, July 28–30 2015 p3452

    [19]

    Tran S J D, Teuscher C 2017 IEEE/ACM International Symposium on Nanoscale Architectures (NANOARCH) Newport, RI, July 25−26 2017 p115

    [20]

    Wang Z, Rao M, Han J W, Zhang J, Lin P, Li Y, Li C, Song W, Asapu S, Midya R, Zhuo Y, Jiang H, Yoon J H, Upadhyay N K, Joshi S, Hu M, Strachan J P, Barnell M, Wu Q, Wu H, Qiu Q, Williams R S, Xia Q, Yang J J 2018 Nat Commun. 9 3208Google Scholar

    [21]

    Chen Y, Zhang J, Zhang Y, Zhang R, Kimura M, Nakashima Y 2019 17th IEEE International New Circuits and Systems Conference (NEWCAS) Munich, Germany, June 23−26 2019 p1

    [22]

    Tran S J D, Teuscher C 2019 IEEE International Conference on Rebooting Computing (ICRC) San Mateo, CA, Nov. 6−8 2019 p110

    [23]

    L.Chua 2015 Radioengineering 24 319Google Scholar

    [24]

    Bessonov A A, Kirikova M N, Petukhov D I, Allen M, Ryhanen T, Bailey M J 2015 Nat. Mater. 14 199Google Scholar

    [25]

    Goswami S, Rath S P, Thompson D, Hedstrom S, Annamalai M, Pramanick R, Ilic B R, Sarkar S, Hooda S, Nijhuis C A, Martin J, Williams R S, Goswami S, Venkatesan T 2020 Nat. Nanotechnol. 15 380Google Scholar

    [26]

    Lai Q X, Zhang L, Li Z Y, Stickle W F, Williams R S, Chen Y 2009 Appl. Phys. Lett. 95 213503Google Scholar

    [27]

    Liu R X, Dong R X, Qin S C, Yan X L 2020 Org. Electron. 81 105680Google Scholar

    [28]

    Liu S Q, Wu N J, Ignatiev A, Li J R 2006 J. Appl. Phys. 100 056101Google Scholar

    [29]

    Martinez-Rincon J, Di Ventra M, Pershin Y V 2010 Phys. Rev. B. 81 195430Google Scholar

    [30]

    Najem J S, Hasan M S, Williams R S, Weiss R J, Rose G S, Taylor G J, Sarles S A, Collier C P 2019 Nat Commun. 10 3239Google Scholar

    [31]

    Nieminen H, Ermolov V, Nybergh K, Silanto S, Ryhanen T 2002 J. Micromech. Microeng. 12 177Google Scholar

    [32]

    Noh Y J, Baek Y J, Hu Q, Kang C J, Choi Y J, Lee H H, Yoon T S 2015 IEEE Trans. Nanotechnol. 14 798Google Scholar

    [33]

    Park D, Yang P, Kim H J, Beom K, Lee H H, Kang C J, Yoon T S 2018 Appl. Phys. Lett. 113 162102Google Scholar

    [34]

    Román Acevedo W, van den Bosch C A M, Aguirre M H, Acha C, Cavallaro A, Ferreyra C, Sánchez M J, Patrone L, Aguadero A, Rubi D 2020 Appl. Phys. Lett. 116 063502Google Scholar

    [35]

    Salaoru I, Khiat A, Li Q J, Berdan R, Prodromakis T 2013 Appl. Phys. Lett. 103 233513Google Scholar

    [36]

    Slesazeck S, Wylezich H, Mikolajick T 2017 IEEE 8th Latin American Symposium on Circuits & Systems (LASCAS) Bariloche, Argentina, Feb. 20−23 2017 p1

    [37]

    Sun J, Lind E, Maximov I, Xu H Q 2011 IEEE Electron Device Lett. 32 131Google Scholar

    [38]

    Wu S X, Peng H Y, Wu T 2011 Appl. Phys. Lett. 98 093503Google Scholar

    [39]

    Ahmed M G, Cho K, Cho T 2012 13th International Workshop on Cellular Nanoscale Networks and their Applications Turin, Italy, Aug. 29−31 2012 p1

    [40]

    Asapu S, Pershin Y V 2015 IEEE Trans. Electron Devices 62 3678Google Scholar

    [41]

    Biolek D, Biolek Z, Biolkova V 2009 European Conference on Circuit Theory and Design Antalya, Turkey, Aug. 23−27 2009 p249

    [42]

    Biolek D, Biolek Z, Biolkova V 2010 Electron. Lett. 46 520Google Scholar

    [43]

    Biolek D, Biolkova V 2010 Electron. Lett. 46 1428Google Scholar

    [44]

    Biolek D, Biolková V, Kolka Z 2010 IEEE Asia Pacific Conference on Circuits and Systems Kuala Lumpur, Malaysia Dec. 6−9 2010 p800

    [45]

    Flak J, Raantala A, Haatainen T, Prunnila M, Laiho M 2014 14th International Workshop on Cellular Nanoscale Networks and their Applications (CNNA) Notre Dame, IN, USA, July 29−31 2014 p1

    [46]

    Fouda M E, Radwan A G 2012 Electron. Lett. 48 1454Google Scholar

    [47]

    Pershin Y V, Di Ventra M 2010 Electron. Lett. 46 517Google Scholar

    [48]

    Pershin Y V, Di Ventra M 2011 Electron. Lett. 47 243Google Scholar

    [49]

    Romero F J, Morales D P, Godoy A, Ruiz F G, Tienda-Luna I M, Ohata A, Rodriguez N 2019 Int. J. Circ. Theor. App. 47 572Google Scholar

    [50]

    Yu D S, Liang Y, Iu H H C, Chua L O 2014 IEEE Trans. Circuits Syst. II-Express Briefs 61 758Google Scholar

    [51]

    Yu D, Zhao X, Sun T, Iu H H C, Fernando T 2020 IEEE Trans. Circuits Syst. II-Express Briefs 67 1334Google Scholar

    [52]

    Yu D, Zhou Z, Iu H H C, Fernando T, Hu Y 2016 IEEE Trans. Circuits Syst. II-Express Briefs 63 1101Google Scholar

    [53]

    Yu D S, Liang Y, Chen H, Iu H H C 2013 IEEE Trans. Circuits Syst. II-Express Briefs 60 207Google Scholar

    [54]

    Zheng C Y, Yu D S, Iu H H C, Fernando T, Sun T T, Eshraghian J K, Guo H D 2019 IEEE Trans. Circuits Syst. I-Regul. Pap. 66 4793Google Scholar

    [55]

    Kwon D, Chung I Y 2020 IEEE Electron Device Lett. 41 493Google Scholar

    [56]

    Zhao L, Fan Z, Cheng S L, Hong L Q, Li Y Q, Tian G, Chen D Y, Hou Z P, Qin M H, Zeng M, Lu X B, Zhou G F, Gao X S, Liu J M 2020 Adv Electron Mater 6 1900858Google Scholar

    [57]

    Yamaletdinov R D, Ivakhnenko O V, Sedelnikova O V, Shevchenko S N, Pershin Y V 2018 Sci. Rep. 8 3566Google Scholar

    [58]

    Patel J A, Sandhie Z T, Chowdhury M H 2018 IEEE 61st International Midwest Symposium on Circuits and Systems (MWSCAS) Windsor, Canada, Aug. 5–8 2018 p1130

    [59]

    Salaoru I, Li Q, Khiat A, Prodromakis T 2014 Nanoscale. Res. Lett. 9 552Google Scholar

    [60]

    Cai J W, Li L X, Xu C, Feng Y, Zhong Y N, Xu J L, Gao X, Wang S D 2019 Appl. Phys. Lett. 114 043302Google Scholar

    [61]

    Qian W H, Cheng X F, Zhao Y Y, Zhou J, He J H, Li H, Xu Q F, Li N J, Chen D Y, Lu J M 2019 Adv. Mater. 31 1806424Google Scholar

    [62]

    Yang P, Jun Kim H, Zheng H, Won Beom G, Park J S, Jung Kang C, Yoon T S 2017 Nanotechnology 28 225201Google Scholar

    [63]

    Martinez-Rincon J, Pershin Y V 2011 IEEE Trans. Electron Devices 58 1809Google Scholar

    [64]

    Yang C, Yang N, Yu Y, Li Y, Diez F F 2017 IEEE 17th International Conference on Communication Technology (ICCT) Chengdu, China, Oct. 27–30 2017 p1171

    [65]

    Corinto F, Di Marco M, Forti M, Chua L 2019 IEEE Trans Cybern 50 4758Google Scholar

    [66]

    Cohen G Z, Pershin Y V, Di Ventra M 2012 Phys. Rev. B. 85 165428Google Scholar

    [67]

    Mcculloch W S, Pitts W 1943 Bull. Math. Biol. 5 115

    [68]

    Hodgkin A L, Huxley A F 1989 Bull. Math. Biol. 52 25

    [69]

    Pershin Y V, Di Ventra M 2011 Adv. Phys. 60 145Google Scholar

    [70]

    Rumelhart D E, Hinton G E, Williams R J 1986 Nature 323 533Google Scholar

    [71]

    John H, Anders K, Palmer R G 1991 Phys. Today 44 70

    [72]

    Bi G Q, Poo M M 2001 Annu. Rev. Neurosci. 24 139Google Scholar

  • [1] 黄宇航, 陈理想. 基于未训练神经网络的分数傅里叶变换成像. 物理学报, 2024, 0(0): 0-0. doi: 10.7498/aps.73.20240050
    [2] 马锐垚, 王鑫, 李树, 勇珩, 上官丹骅. 基于神经网络的粒子输运问题高效计算方法. 物理学报, 2024, 73(7): 072802. doi: 10.7498/aps.73.20231661
    [3] 陈开辉, 樊贞, 董帅, 李文杰, 陈奕宏, 田国, 陈德杨, 秦明辉, 曾敏, 陆旭兵, 周国富, 高兴森, 刘俊明. 钙钛矿相界面插层对SrFeOx基忆阻器的性能提升. 物理学报, 2023, 72(9): 097301. doi: 10.7498/aps.72.20221934
    [4] 李瑞, 徐邦林, 周建芳, 姜恩华, 汪秉宏, 袁五届. 一种突触可塑性导致的觉醒-睡眠周期中突触强度变化和神经动力学转变. 物理学报, 2023, 72(24): 248706. doi: 10.7498/aps.72.20231037
    [5] 方波浪, 王建国, 冯国斌. 基于物理信息神经网络的光斑质心计算. 物理学报, 2022, 71(20): 200601. doi: 10.7498/aps.71.20220670
    [6] 李靖, 孙昊. 识别Z玻色子喷注的卷积神经网络方法. 物理学报, 2021, 70(6): 061301. doi: 10.7498/aps.70.20201557
    [7] 朱玮, 刘兰, 文常保, 李杰. 双层结构突触仿生忆阻器的时空信息传递及稳定性. 物理学报, 2021, 70(17): 178504. doi: 10.7498/aps.70.20210274
    [8] 孙立望, 李洪, 汪鹏君, 高和蓓, 罗孟波. 利用神经网络识别高分子链在表面的吸附相变. 物理学报, 2019, 68(20): 200701. doi: 10.7498/aps.68.20190643
    [9] 顾梅园, 刘敬彪, 王光义, 梁燕, 李付鹏. 忆容器多谐振荡器及其实验. 物理学报, 2019, 68(22): 228401. doi: 10.7498/aps.68.20190849
    [10] 魏德志, 陈福集, 郑小雪. 基于混沌理论和改进径向基函数神经网络的网络舆情预测方法. 物理学报, 2015, 64(11): 110503. doi: 10.7498/aps.64.110503
    [11] 李欢, 王友国. 一类非线性神经网络中噪声改善信息传输. 物理学报, 2014, 63(12): 120506. doi: 10.7498/aps.63.120506
    [12] 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
    [13] 陈铁明, 蒋融融. 混沌映射和神经网络互扰的新型复合流密码. 物理学报, 2013, 62(4): 040301. doi: 10.7498/aps.62.040301
    [14] 李华青, 廖晓峰, 黄宏宇. 基于神经网络和滑模控制的不确定混沌系统同步. 物理学报, 2011, 60(2): 020512. doi: 10.7498/aps.60.020512
    [15] 赵海全, 张家树. 混沌通信系统中非线性信道的自适应组合神经网络均衡. 物理学报, 2008, 57(7): 3996-4006. doi: 10.7498/aps.57.3996
    [16] 王永生, 孙 瑾, 王昌金, 范洪达. 变参数混沌时间序列的神经网络预测研究. 物理学报, 2008, 57(10): 6120-6131. doi: 10.7498/aps.57.6120
    [17] 王瑞敏, 赵 鸿. 神经元传输函数对人工神经网络动力学特性的影响. 物理学报, 2007, 56(2): 730-739. doi: 10.7498/aps.56.730
    [18] 王耀南, 谭 文. 混沌系统的遗传神经网络控制. 物理学报, 2003, 52(11): 2723-2728. doi: 10.7498/aps.52.2723
    [19] 谭文, 王耀南, 刘祖润, 周少武. 非线性系统混沌运动的神经网络控制. 物理学报, 2002, 51(11): 2463-2466. doi: 10.7498/aps.51.2463
    [20] 神经网络的自适应删剪学习算法及其应用. 物理学报, 2001, 50(4): 674-681. doi: 10.7498/aps.50.674
计量
  • 文章访问数:  9090
  • PDF下载量:  453
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-10-02
  • 修回日期:  2020-11-06
  • 上网日期:  2021-03-30
  • 刊出日期:  2021-04-05

/

返回文章
返回