Dynamic Event-triggered Control And Attack Estimation Of Cyber-physical Systems Under Cyber Attacks

As a large-scale,distributed,heterogeneous,multi-dimensional and complex intelligent system,integrating computing,communication and control,cyber-physical system(CPS)realizes the coordinated operation of computing resources and physical resources.At present,it is widely used in various engineering fields.In this context,network attacks become a challenge for CPS security now.Therefore,it is of great significance to solve the network security problem of CPS under network attacks.Based on the analysis of the above situation,this paper carries out a series of research on how to resist the network attacks in CPS,and achieved the following research results.Firstly,the analysis and design of resilient triggering controller for multi-area load frequency control(LFC)system under DOS attacks are studied.Considering that the LFC system is disturbed by aperiodic DoS attack,the average dwell time(ADT)method is used to constrain the upper bound of the frequency and duration of DoS attacks,and the dynamic event-triggering method is introduced to reduce the system communication load.Then,the LFC system under DOS attacks is modeled as a switching time-delay system.Then,combined with Lyapunov control theory and switched system theory,the stability and weighting H_∞performance of the system are analyzed.On this basis,a design method of resilient triggering controller is proposed based on linear matrix inequality(LMIs)technology.Finally,the effectiveness and superiority of this method are verified by simulation experiments.Secondly,the secure frequency control of multi-area hybrid power system with wind power under DoS attacks is studied.Considering that the control channels of thermal generator and wind farm are disturbed by DoS attacks,the hybrid power system is modeled as a switched time-delay system.Based on the above model,by constructing Lyapunov Krasovskii functional and reasonably using LMIs technology,the conditions for exponential stability of the system and the design parameters of the controller are derived.Finally,the effectiveness of the method is verified by simulation experiments.Finally,the problem of attacks estimation of CPS under the interference of false data injection(FDI)attacks is studied.Considering that the system is attacked by both sensors and actuators,an unknown input observer method is proposed to estimate the FDI attacks on sensors and actuators.Firstly,the necessary and sufficient conditions for the existence of state observer and attacks estimator are given.Then,under a specific performance level,the l_∞stability of system state and attacks estimation error are guaranteed.According to LMIs technology,the design of state observer and attacks estimator are realized.Finally,the effectiveness of the method is verified by numerical simulation.