Memristive neural network circuit with fault tolerance for character recognition

doi:10.1088/1674-1056/ae3f94

中国物理B ›› 2026, Vol. 35 ›› Issue (6): 60702-060702.doi: 10.1088/1674-1056/ae3f94

Memristive neural network circuit with fault tolerance for character recognition

Mei Guo(郭梅)¹, Jikang Liu(刘继康)¹, and Jingzhi Xu(徐景芝)^2,†

1 College of Electrical Engineering and Automation, Shandong University of Science and Technology, Qingdao 266590, China;
2 Department of Science and Technology Management, Shandong University of Science and Technology, Qingdao 266590, China

收稿日期:2025-12-17 修回日期:2026-01-30 接受日期:2026-01-30 发布日期:2026-06-15
通讯作者: Jingzhi Xu E-mail:skd994298@sdust.edu.cn
基金资助:
Project supported by the Natural Science Foundation of Shandong Province (Grant No. ZR2022MF225) and the National Natural Science Foundation of China (Grant Nos. 62176143 and 62371275).

Memristive neural network circuit with fault tolerance for character recognition

Mei Guo(郭梅)¹, Jikang Liu(刘继康)¹, and Jingzhi Xu(徐景芝)^2,†

1 College of Electrical Engineering and Automation, Shandong University of Science and Technology, Qingdao 266590, China;
2 Department of Science and Technology Management, Shandong University of Science and Technology, Qingdao 266590, China

Received:2025-12-17 Revised:2026-01-30 Accepted:2026-01-30 Published:2026-06-15
Contact: Jingzhi Xu E-mail:skd994298@sdust.edu.cn
Supported by:
Project supported by the Natural Science Foundation of Shandong Province (Grant No. ZR2022MF225) and the National Natural Science Foundation of China (Grant Nos. 62176143 and 62371275).

摘要/Abstract

摘要： Memristor-based neural networks are one of the most promising approaches for the hardware implementation of artificial neural networks. In this paper, a memristor-based neural network circuit based on a one-memristor-one-resistor (1M1R) synaptic array structure is designed for character recognition. Compared with other memristive synaptic arrays, the 1M1R structure can reduce the number of memristors used. However, memristors may malfunction due to fabrication defects and the influence of external factors, resulting in a decrease in the accuracy of the circuit's character recognition, and a suitable solution needs to be found to improve the stability and durability of the circuit. Therefore, in this paper, a fault-tolerant module with feedback adjustment capability is designed in the memristive neural network circuit that can re-adjust the weights of the memristors through in-situ training to solve multiple faults in the memristive neural network. The effect of fault tolerance is verified by character recognition. The experimental results show that the designed memristive neural network circuit can accurately realize character recognition, and the designed fault-tolerant circuit can well tolerate multiple faults, ensuring stable operation of the circuit under fault conditions.

关键词: memristor, neural network circuit, character recognition, fault-tolerant, feedback

Abstract: Memristor-based neural networks are one of the most promising approaches for the hardware implementation of artificial neural networks. In this paper, a memristor-based neural network circuit based on a one-memristor-one-resistor (1M1R) synaptic array structure is designed for character recognition. Compared with other memristive synaptic arrays, the 1M1R structure can reduce the number of memristors used. However, memristors may malfunction due to fabrication defects and the influence of external factors, resulting in a decrease in the accuracy of the circuit's character recognition, and a suitable solution needs to be found to improve the stability and durability of the circuit. Therefore, in this paper, a fault-tolerant module with feedback adjustment capability is designed in the memristive neural network circuit that can re-adjust the weights of the memristors through in-situ training to solve multiple faults in the memristive neural network. The effect of fault tolerance is verified by character recognition. The experimental results show that the designed memristive neural network circuit can accurately realize character recognition, and the designed fault-tolerant circuit can well tolerate multiple faults, ensuring stable operation of the circuit under fault conditions.

Key words: memristor, neural network circuit, character recognition, fault-tolerant, feedback

中图分类号: (Neural networks, fuzzy logic, artificial intelligence)

07.05.Mh

45.30.+s (General linear dynamical systems)

Mei Guo(郭梅), Jikang Liu(刘继康), and Jingzhi Xu(徐景芝). Memristive neural network circuit with fault tolerance for character recognition[J]. 中国物理B, 2026, 35(6): 60702-060702.

Mei Guo(郭梅), Jikang Liu(刘继康), and Jingzhi Xu(徐景芝). Memristive neural network circuit with fault tolerance for character recognition[J]. Chin. Phys. B, 2026, 35(6): 60702-060702.

参考文献

[1] Wang X Y, Zhou P F, Eshraghian J K, Lin C Y, Iu H H C, Chang T C and Kang S M 2021 IEEE Trans. Circuits Syst. I-Regul. Pap. 68 264
[2] Rao M Y, Tang H, Wu J B, Song W H, Zhang M, Yin W B, Zhuo Y, Kiani F, Chen B J M, Jiang X Q, Liu H F, Chen H Y, Midya Y, Ye F, Jiang H, Wang Z R, Wu M C, Hu M, Wang H, Xia Q F, Ge N, Li J and Yang J J 2023 Nature 615 823
[3] Chua L 1971 IEEE Trans. Circuit Theory 18 507
[4] Strukov D B, Snider G S, Stewart D R and Williams R S 2008 Nature 453 80
[5] Karimi A and Rezai A 2019 Int. J. Circuit Theory Appl. 47 1933
[6] Rziga F O, Mbarek K, Ghedira S and Besbes K 2019 J. Comput. Electron. 18 1055
[7] Mozaffari S N, Tragoudas S and Haniotakis T 2017 IEEE Trans. 36 1018
[8] Singh A 2020 IETE J. Res. 66 182
[9] Papandroulidakis G, Serb A, Khiat A and Merrett G V 2019 IEEE Trans. Circuits Syst. I-Regul. Pap. 66 3041
[10] Gul F 2019 IEEE Electron Device Lett. 40 643
[11] Guo M, Zhang X W, Dou G and Iu H H C 2025 IEEE Trans. Circuits Syst. I-Regul. Pap. 72 7120
[12] Lee J Y, Han J Y, Kang B X, Hong J Y, Lee C Z and Jeon I 2025 Adv. Mater. 37 2413916
[13] Guo M, Zhang X W, Guo W H, Dou G, Chen D, Wang L H, Iu H H C 2025 IEEE Trans. Circuits Syst. I-Regul.
[14] Sun J W, Zhai Y, Liu P and Wang Y F 2025 IEEE Trans. Neural Netw. Learn. Syst. 36 3618
[15] Guo M, Sun J H, Dou G and Iu H H C 2025 IEEE Trans. Circuits Syst. I-Regul. Pap. 72 5998
[16] Dou G, Li D G, Guo M and Iu H H C 2025 IEEE Trans. Circuits Syst. I-Regul. Pap. 72 2566
[17] Wang Y F, Tao K F, Wang Z C and Sun J W 2025 IEEE Trans. Ind. Inform.
[18] Sun J W, Gao P L, Liu P and Wang Y F 2025 IEEE Trans. Ind. Inform.
[19] Guo M, Kong L T, Dou G and Iu H H C 2024 IEEE Trans. Circuits Syst. I-Regul. Pap. 71 4676
[20] Guo M, Zhang D Y, Guo W H, Dou G and Sun J W 2024 IEEE Trans. Cogn. Dev. Syst. 17 155
[21] Guo M, Zheng C G, Dou G and Iu H H C A 2026 IEEE Trans. Circuits Syst. I-Regul. Pap. 73 850
[22] Lei Z, Guo Y T and Ma J 2025 Nonlinear Dyn. 113 29999
[23] Wang B C, Guo Y T, Jia J N and Ma J 2025 Euro. Phys. J. Plus 140 706
[24] Ma J 2025 Nonlinear Dyn. 113 25365
[25] Lu J L, Gao J, Cai J Y, Chen L X, Zhang G Y and Xu Q 2025 Chin. Phys. B 34 018703
[26] Yan X, Wu F Q, Yao W, Zhou P and Liu Y Z 2024 Chin. Phys. B 33 028705
[27] Lu J Y, Wang G Y, Xu Y, Ren G D and Ma J 2024 Chin. Phys. B 33 048701
[28] Li J, Xin Y, Sun B, Gu D S, Liao C R, Hu X F, Wang L D, Duan S K and Zhou G D 2025 J. Semicond. 46 012604
[29] Xiao H, Zhou Y, Gao T T, Duan S K, Chen G R and Hu X F 2023 IEEE J. Emerg. Sel. Top. Circuits Syst. 13 344
[30] Zhao H, Liu Z W, Tang J S, Gao B, Qin Q, Li J M, Zhou Y, Yao P, Xi Y, Lin Y D and Wu H Q 2023 Nat. Commun. 14 2276
[31] Kim N, Oh J, Kim S, Cha J H, Choi J, Im S G, Choi S Y and Jang B C 2024 Adv. Funct. Mater. 34 2305136
[32] Wu X L, Dang B J, Wang H, Wu X L and Yang Y C 2021 Adv. Intell. Syst. 4 2100151
[33] Yang L, Zeng Z G and Shi X M 2019 Neurocomputing 363 114
[34] Chen C Y, Shih H C, Wu C W, Lin C H, Chiu P F, Sheu S S and Chen F T 2015 IEEE Trans. Comput. 64 180
[35] Sun X Y and Yu S M 2019 IEEE J. Emerg. Sel. Top. Circuits Syst. 9 570
[36] Zhang B G, Uysal N, Fan D L and Ewetz R 2020 IEEE Trans. Comput.- Aided Des. Integr. Circuits Syst. 39 2448
[37] Liu P, You Z Q, Wu J G, Liu B S, Han Y H and Chakrabarty K 2021 IEEE Trans. Circuits Syst. 68 4444
[38] Chen C Y and Chakrabarty K 2022 IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst. 41 2301
[39] Morgul M C, Frontini L, Tunali O, Anghel L, Ciriani V, Vatajelu E I, Moritz C A, StanMR, Alexandrescu D and AltunM2021 IEEE Trans. Nanotechnol. 20 39
[40] Bi Y T, Xu Q, Geng H, Chen S and Kang Y 2023 IEEE Trans. Circuits Syst. II-Express Briefs 70 2221
[41] Wang Z W, Yin M H, Zhang T, Cai Y M, Wang Y Y, Yang Y C and Huang R 2016 Nanoscale 8 14015
[42] Chen A 2016 Solid-State Electron. 125 25
[43] Berdan R, Prodromakis T and Toumazou C 2012 Electron. Lett. 48 1105
[44] Ghofrani A, Lastras-MontanoMA, Gaba S, Payvand M, LuW, Theogarajan L and Cheng K T 2015 J. Emerg. Technol. Comput. Syst. 12 1
[45] Guo M, Zhu Y L, Liu R Y, Zhao K X and Dou G 2022 Neurocomputing 472 12
[46] Zhang X R, Wang X Y, Ge Z Y, Li Z L, Wu M Y and Borah S 2022 Micromachines 13 2074
[47] Quan C H, FoudaME, Lee S, Jung G J, Lee J, Eltawil A E and Kurdahi F 2023 IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst. 42 2174
[48] Reddy M A, Pandeeswari R and Ko S B 2023 IEEE Trans. Circuits Syst. II-Express Briefs 70 959
[49] Shi F, Wang K H, Cao Y H and Mou J A 2026 J. Emerg. Technol. Comput. Syst. 195 108921

Memristive neural network circuit with fault tolerance for character recognition

Memristive neural network circuit with fault tolerance for character recognition

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	Shaobo He(贺少波), Jiawei Xiao(肖佳伟), Qilai Chen(陈祺来), and Huihai Wang(王会海). Fixed points as regulatory hubs in discrete memristive neural networks: An analysis of the FitzHugh-Nagumo model[J]. 中国物理B, 2026, 35(6): 60502-060502.
[2]	Jieyu Lu(鲁婕妤), Jiapeng Ouyang(欧阳佳鹏), Xue Zhao(赵雪), and Minglin Ma(马铭磷). Dynamical behavior analysis for small-world scale-free neural networks[J]. 中国物理B, 2026, 35(6): 68708-068708.
[3]	Ke Meng(孟珂), Yifan Bu(卜一帆), Yinghong Cao(曹颖鸿), Suo Gao(高锁), Qi Li(李琦), Chunpeng Wang(王春鹏), and Jun Mou(牟俊). Chaotic bursting and burst synchronization in a discrete dual-Rulkov neural network with memristive synaptic coupling[J]. 中国物理B, 2026, 35(6): 60501-060501.
[4]	Ying Xu(徐莹). Synergistic delay and strength tuning for logical operations in a chaotic neuron[J]. 中国物理B, 2026, 35(6): 60201-060201.
[5]	Zhicheng Tian(田桎成), Peipei Jin(靳培培), Shutong Liu(刘姝彤), Meiyuan Gu(顾梅园), Long Chen(陈龙), and Guangyi Wang(王光义). Symmetrical Turing instability in Chua corsage memristor siblings-based two-cell network[J]. 中国物理B, 2026, 35(6): 60511-060511.
[6]	Wei Yao(姚卫), Xiangyun Huang(黄翔云), Jianhua Xiao(肖捡花), Fei Yu(余飞), and Yichuang Sun(孙义闯). A visually meaningful medical image encryption scheme based on image steganography and memristive Hopfield neural networks[J]. 中国物理B, 2026, 35(6): 68702-068702.
[7]	Peng Qin(秦鹏), Tieqiao Liu(刘铁桥), Qiuzhen Wan(万求真), Rou Zhou(周柔), and Huaimin Xiang(向怀民). Environmental-gradient emotional memory memristive neural network circuit with TAP cell regulatory mechanism[J]. 中国物理B, 2026, 35(6): 68709-068709.
[8]	Yueqi Song(宋钥淇), Xiaozhou He(何晓舟), Xianying Xu(徐宪莹), Yinghong Cao(曹颖鸿), Santo Banerjee, Suo Gao(高锁), and Jun Mou(牟俊). Echolocation-inspired memristive behavioral decision circuit[J]. 中国物理B, 2026, 35(6): 60701-060701.
[9]	Qijian Wu(吴奇键), Peipei Jin(靳培培), Xiameng Wu(吴夏萌), Meiyuan Gu(顾梅园), Wei Zhou(周玮), Yujiao Dong(董玉姣), Yan Liang(梁燕), and Long Chen(陈龙). Frequency-tunable spiking dynamics of Chua corsage memristors[J]. 中国物理B, 2026, 35(4): 48703-048703.
[10]	Yuxia Li(李玉霞), Xintong Yue(岳新同), Hui Chang(常辉), Baoxing Han(韩宝兴), and Yan Zhang(张燕). Spiking-bursting alternating chaos mediated by a locally active memristor[J]. 中国物理B, 2026, 35(3): 30504-030504.
[11]	Xiao Huang(黄潇), Peiwen Tong(童霈文), Qingjiang Li(李清江), Tuo Ma(马拓), Shuo Han(韩硕), Wei Wang(王伟), and Yi Sun(孙毅). Memristor-based analog noise correction for infrared sensors[J]. 中国物理B, 2026, 35(2): 28501-028501.
[12]	Yichen Bi(毕毅晨), Jun Mou(牟俊), Herbert Ho-Ching Iu, Nanrun Zhou(周南润), Santo Banerjee, and Suo Gao(高锁). Dynamics analysis and DSP implementation of the Rulkov neuron model with memristive synaptic crosstalk[J]. 中国物理B, 2026, 35(1): 10504-010504.
[13]	Jingru Sun(孙晶茹), Xiaosong Li(李晓崧), Yichuang Sun(孙义闯), Zining Xiong(熊子宁), and Jiqi He(何计奇). An artificial synapse capable of regulating signal transmission speed in a neuromorphic network[J]. 中国物理B, 2026, 35(1): 10501-010501.
[14]	Shuang Zhao(赵双), Yunzhen Zhang(张云贞), Xiangjun Chen(陈湘军), Bin Gao(高彬), and Chengjie Chen(陈成杰). Dynamic analysis and DNA coding-based image encryption of memristor synapse-coupled hyperchaotic IN-HNN network[J]. 中国物理B, 2026, 35(1): 10502-010502.
[15]	Zhao-Bin Zeng(曾昭彬), Ya-Hui Sun(孙亚辉), and Yang Liu(刘洋). Probabilistic distribution and stochastic P-bifurcation of a nonlinear energy-regenerative suspension system with time-delayed feedback control[J]. 中国物理B, 2026, 35(1): 10201-010201.