混合网络攻击下模糊 Markov 跳变时滞神经网络异步耗散故障诊断滤波

Asynchronous dissipative fault diagnosis filtering for fuzzy Markov jump time-delay neural networks under hybrid cyber attacks

摘要

研究了混合网络攻击下模糊 Markov 跳变时滞神经网络的异步故障诊断滤波。引入动态事件触发机制有效减少信息传输过程中网络带宽拥塞, 缓解通信负担。由于网络资源的开放性, 信息在传输中易遭受网络攻击, 采用服从伯努利分布的随机变量描述系统受到欺骗攻击和拒绝服务攻击。由于系统时滞及外界干扰等因素造成滤波器难以完美获得系统模态信息, 利用隐 Markov 模型刻画系统和滤波器模态之间的异步行为。根据 T-S 模糊法和随机分析法, 首次建立混合网络攻击下离散时间模糊 Markov 跳变时滞神经网络故障诊断系统, 并设计故障诊断滤波器来生成残差信号, 判断是否有故障发生。利用 Lyapunov 函数法和迭代等方法, 得到故障诊断滤波误差系统随机均方稳定且满足严格耗散性能的充分性判据, 在此基础上设计异步故障滤波器增益矩阵。最后通过数值仿真验证了所给结果的正确性。

Abstract

This paper investigates the asynchronous fault diagnosis filtering for fuzzy Markov jump time-delay neural networks under hybrid cyber attacks. The dynamic event-triggered mechanism is introduced to effectively reduce network bandwidth congestion during information transmission, thereby alleviating communication burdens. To address the susceptibility of information to cyber attacks during transmission due to the openness of network resources, Bernoulli-distributed stochastic variables are employed to characterize the system’s exposure to deception attacks and denial-of-service (DoS) attacks. Because the existence of time delays and external disturbances make it difficult for the filter to perfectly obtain system model information, the hidden Markov model (HMM) is established to characterize the asynchronous behavior between the system mode and the filter mode. Based on the T-S fuzzy method and stochastic analysis, the fault diagnosis system for discrete-time fuzzy Markov jump time-delay neural networks under hybrid cyber attacks is first established and the fault diagnosis filter is designed to generate residual signals, which are utilized to determine the occurrence of faults. By employing the Lyapunov function method and iterative techniques, sufficient criteria are derived to ensure that the fault diagnosis error system is stochastically mean-square stable and satisfies the strict dissipative performance. On this basis, the gain matrix of the asynchronous fault diagnosis filter is designed. Finally, the correctness of the obtained results is verified by a numerical example.

关键词

神经网络 / Markov 跳变系统 / 网络攻击 / 故障诊断 / 耗散滤波

Key words

neural networks / Markov jump systems / cyber attacks / fault diagnosis / dissipative filtering

参考文献

[1] MA S W , ZHANG X F , JIA C M , et al. Image and video compression with neural networks: A review[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2019, 30(6): 1683-1698.

[2] ZHUANG H P , LIN Z P , YANG Y Y , et al. An analytic formulation of convolutional neural network learning for pattern recognition[J/OL]. Information Sciences, 2025, 686.DOI: 10.1016/j.ins.2024.121317.

[3] LIN C M , BOLDBAATAR E A . Autolanding control using recurrent wavelet Elman neural network[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2015, 45(9): 1281-1291.

[4] LIN W J , TAN G Q , WANG Q G , et al. Fault—tolerant state estimation for Markov jump neural networks with time—varying delays[J]. IEEE Transactions on Circuits and Systems II: Express Briefs, 2023, 71(4): 2114-2118.

[5] XUE M , YAN H C , WANG M , et al. LSTM—based intelligent fault detection for fuzzy Markov jump systems and its application to tunnel diode circuits[J]. IEEE Transactions on Circuits and Systems II: Express Briefs, 2021, 69(3): 1099-1103.

[6] ZHANG Y G, XIA J W, XIA H, et al. Asynchronous l ₂—l _∞ filtering for discrete—time fuzzy Markov jump neural networks with unreliable communication links [J]. Neural Processing Letters, 2020, 52(3): 2069-2088.

[7] CHENG J , XU J M , YAN H C , et al. Neural network—based sliding mode control for semi—Markov jump systems with singular perturbation[J]. IEEE Transactions on Cybernetics, 2024, 55: 259-268.

[8] 钟麦英, 刘浩, 朱晓强, 等 . 基于事件触发技术的故障诊断技术研究综述[J]. 山东科技大学学报 (自然科学版), 2021, 40(2): 95-102.

[9] ZHONG Maiying , LIU Hao , ZHU Xiaoqiang , et al. A survey on recent studies of event—triggered fault diagnosis technology[J]. Journal of Shandong University of Science and Technology (Natural Science), 2021, 40(2): 95-102.

[10] WANG Y F , LI Z X , WANG P L , et al. Robust fault detection for networked Markov jump systems with random time—delay[J/OL]. Mathematical Problems in Engineering, 2017, 1(1).DOI: 10.1155/2017/2608140.

[11] SAKTHIVEL R , DIVYA H , PARIVALLAL A , et al. Quantized fault detection filter design for networked control system with Markov jump parameters[J]. Circuits, Systems, and Signal Processing: CSSP, 2021, 40(10): 4741-4758.

[12] WU Z G , SHI P , SHU Z , et al. Passivity—based asynchronous control for Markov jump systems[J]. IEEE Transactions on Automatic Control, 2016, 62(4): 2020-2025.

[13] SHEN Y , WU Z G , SHI P , et al. Asynchronous filtering for Markov jump neural networks with quantized outputs[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2018, 49(2): 433-443.

[14] DONG S L, WU Z G, SHI P, et al. Networked fault detection for Markov Jump nonlinear systems[J]. IEEE Transactions on Fuzzy Systems, 2018, 26(6): 3368-3378.

[15] YUE D , TIAN E G , HAN Q L . A delay system method for designing event—triggered controllers of networked control systems, IEEE Transactions on Automation Control, 2012, 58(2): 475-481.

[16] PAN Y N , YANG G H . Event—triggered fault detection filter design for nonlinear networked systems[J]. IEEE Transactions on Systems, Man and Cybernetics: Systems, 2018, 48(11): 1851-1862.

[17] PAN Y N , YANG G H . Event—driven fault detection for discrete—time interval type—2 fuzzy systems[J]. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2019, 51(8): 4959-4968.

[18] 汪威辰, 苏磊, 费习宏, 等 . 混合网络攻击下马尔可夫跳变神经网络的安全容错同步控制[J]. 安徽工业大学学报 (自然科学版), 2024, 41(5): 490-498.

[19] WANG Weichen , SU Lei , FEI Xihong , et al. Secure fault—tolerant synchronous control of Markov jumping neural networks under hybrid cyber attacks[J]. Journal of Anhui University of Technology (Natural Science), 2024, 41(5): 490-498.

[20] YAO L, HUANG X. Memory—based adaptive event—triggered secure control of Markovian jump neural networks suffering from deception attack[J]. Science China Technological Sciences, 2023, 66(2): 468-480.

[21] SHANG S T , ZHANG L C , PAN Y N . Event—based finite—time H _∞ security control for networked control systems with deception attacks [J]. International Journal of Fuzzy Systems, 2025, 27(3): 853-865.

[22] YAO L , HUANG X , WANG Z , et al. A multi—sensor—based switching event—triggered mechanism for synchronization control of Markovian jump neural networks under DoS attacks[J]. IEEE Transactions on Information Forensics and Security, 2024, 19: 7548-7559.

[23] DENG Y H , LU H Q , ZHOU W N . Security event—triggered control for Markovian jump neural networks against actuator saturation and hybrid cyber attacks[J]. Journal of the Franklin Institute, 2021, 358(14): 7096-7118.

基金资助

国家自然科学基金项目(62373229)