基于观测器的周期拒绝服务攻击网络化系统动态事件触发控制

Observer-based dynamic event triggering control for networked systems with periodic denial-of-service attack

中文关键词  网络化系统  DoS攻击  动态事件触发  切换系统  观测器

英文关键词  networked systems  denial-of-service attack  dynamic event triggering  switching systems  observer

中文摘要

      针对具有周期拒绝服务(DoS)攻击的网络化系统, 设计一种基于观测器的具有动态事件触发策略的控制器. 首先, 通过DoS攻击对网络化系统的影响建立了DoS攻击模型, 采用切换系统的方法, 将具有DoS攻击的网络化系统 分为DoS攻击活跃子系统和DoS攻击休眠子系统. 对不可测的系统状态设计状态观测器, 通过在静态事件触发中引 入一个内部动态参数, 设计了动态事件触发机制. 利用动态事件触发序列设计基于观测器状态的控制器, 同时, 为保 证系统在DoS攻击活跃期间子系统的稳定性, 将当前最后一次成功触发状态作为DoS攻击活跃的控制信号, 通过引 入两个辅助函数得到闭环增广系统. 其次, 依据Lyapunov稳定性理论、Jensen’s不等式和Schur补引理, 得到两个切换 子系统稳定充分条件, 在此基础上引入DoS攻击限制得到具有LMI形式的切换系统渐近稳定的充分条件, 同时给出 了状态观测器、控制器和动态事件触发协同设计的参数. 最后, 通过一个数值例子验证所提方法的有效性.

英文摘要

      For networked systems of periodic denial-of-service (DoS) attack, we design a kind of controller with dynamic event triggering strategy based on observer. First of all, the DoS attack model is established through the influence of DoS attack on the networked systems. The networked systems with DoS attack are divided into DoS attack active subsystem and DoS attack dormant subsystem by switching systems method. A state observer is designed for the system with unmeasurable state. By introducing an internal dynamic parameter into the static event triggering, a dynamic event triggering mechanism is designed. A controller based on observer state is designed for sequences of dynamic event triggering. At the same time, in order to ensure stability of the subsystem during the active DoS attack, taking the last successful trigger state as control signal of the active DoS attack, the closed-loop augmented system is obtained by introducing two auxiliary functions. Secondly, two sufficient conditions for stability of switching subsystems are obtained based on Lyapunov stability theory, Jensen’s inequality and Schur complement lemma. On this basis, sufficient conditions for asymptotical stable of switching systems with LMI form are obtained by introducing DoS attack restrictions. Meanwhile, the co-design parameters of state observer, controller and dynamic event triggering mechanism are given. Finally, a numerical example is given to verify the effectiveness of the proposed method.