DoS攻击下网络化控制系统基于观测器的记忆型事件触发预测控制

本文研究了DoS攻击下网络化控制系统记忆型事件触发预测补偿控制问题. 首先, 由于网络带宽资源有限和系统状态不完全可观测性, 引入了记忆型事件触发函数, 为观测器提供离散事件触发传输方案. 然后, 分析了网络传输通道上发生的DoS攻击. 结合上述记忆型事件触发方案, 在控制节点设计一类新颖的预测控制算法, 节省网络带宽资源并主动补偿DoS攻击. 同时, 建立了基于观测器的记忆型事件触发预测控制的闭环系统, 并且分析稳定性.通过线性矩阵不等式(LMI)和Lyapunov稳定性理论, 建立了控制器、观测器和记忆型事件触发矩阵的联合设计方案,并验证了该方案的可行性. 仿真结果表明, 该方案结合记忆型事件触发机制可以有效补偿DoS攻击, 节约网络带宽资源.

Observer memory-based event-triggered predictive control for networked control systems under DoS attacks

In this paper, we study the memory-based event-triggered predictive control of networked control systems(NCSs) under DoS attacks. First, due to the limited network bandwidth resources and incomplete observability of thesystem state, a memory-based event-triggered function is introduced to provide a discrete event-triggered transmissionscheme for the observer. Then, the DoS attacks occurring on the network transmission channel are analyzed. Combiningthe above memory-based event-triggering scheme, a novel class of predictive control algorithms is designed at the controlnode to save the network bandwidth resources and actively compensate for DoS attacks. Meanwhile, a closed-loop systemof observer memory-based event-triggered predictive control is established and analyzed for stability. The design schemeof controller, observer and memory-based event-triggered matrix is established by linear matrix inequality (LMI) andLyapunov stability theory, and the feasibility of the scheme is analyzed. The simulation results show that the schemecombined with the memory-based event-triggered mechanism can effectively compensate for the DoS attacks and save thenetwork bandwidth resources.