Research On Secure Control Of Network Systems Under Denial-of-service Attacks

The birth of 5G technology has led to the development of various Internet of things technologies,making the era of "everything connected" accelerate.The concept of Cyber-Physical Systems was born under this opportunity."Cyber-Physical Systems" means integrating information,communication,and calculation into actual physical engineering systems and realizing the highly efficient,sustainable,and flexible application of energy,transportation,water conservancy,and health systems.In recent years,a series of key technologies around the information physics system have been developed vigorously,such as communication,control,robot,remote operation,etc.In the field of automatic control,the application and development of Cyber-Physical Systems have made profound changes in control theory and application technology,which has changed from closed,closed-loop,and wired systems to open wireless complex interconnected systems.The embedded Internet of things communication and information transmission technology makes the security of the control system be significantly challenged.In fact,over decades,network attacks have been frequent globally,which has posed a significant threat to the development of power systems,intelligent transportation,economy,and industrial production.Therefore,using multi-domain technology to cross-fusion to ensure the safe and stable operation of Cyber-Physical Systems is of great significance to the development of the national output.In this paper,the security control of networked systems under DOS attacks is studied.The security control of interconnected nonlinear large-scale systems is studied by combining adaptive neural networks,adaptive fuzzy systems,average dwell time descriptions,improved counter step method,and Lyapunov Second Stability Theorem.The main works and corresponding safety control design methods of the information physics system are as follows:(1)For a class of interconnected nonlinear large-scale systems with strict feedback form under Denial of Service attacks,firstly,the adaptive neural network system is used to approximate the unknown nonlinear function in the design,that is,the system changes.Then,the adaptive neural network observer deals with the blocking effect of denial of service attack on the channel from the sensor to controller in the control system.Furthermore,the improved backstepping method introduces the switching controller,and the Lyapunov function is used to design the system's controller.The system's stability is analyzed considering the two states of attack activation and silence,and the iterative algorithm is used to reduce the difficulty of solving the observer.Furthermore,Lyapunov's second stability theorem is used to analyze the stability,and the restricted description of DoS attack behavior satisfying the stability condition is given.Finally,the stability of the designed controller is analyzed and verified for the third-order interconnected power network system.(2)For a class of interconnected nonlinear large-scale systems with pure feedback form under Denial of Service attacks,firstly,the system is transformed mathematically to separate the corresponding state and control signals from the state equations.Furthermore,the Butterworth low-pass filter and adaptive fuzzy system are introduced to approximate the system to eliminate the influence of unknown nonlinear dynamics.Secondly,a fuzzy observer with switching gain is designed to observe the state and deal with state transmission blocking.The controller design and stability analysis are then carried out using Lyapunov's second stability theorem and dynamic surface design method.The maximum DoS attack description that the control strategy can tolerate is obtained.Finally,the effectiveness of the safety control design scheme is verified by using MATLAB simulation tool.(3)For a class of interconnected nonlinear large-scale systems with a strict feedback form and unknown control direction under Denial of Service attacks,an adaptive neural network is firstly used to fit and approximate the nonlinear function in the state equations.The problem of unknown control direction is considered,and assumptions are proposed to restrict it.Secondly,combined with the Lyapunov function,a switching form of the adaptive fuzzy observer is designed to observe the unmeasurable state caused by the attack.Based on that,the improved dynamic surface design method combined with the Nussbaum function and first-order filter is used to design the controller and eliminate the influence of the unknown control direction.Through theoretical analysis,the convergence and stability of the fuzzy observer and control system signals are verified,and the limits of DoS attack frequency and duration are given.Finally,two interconnected third-order subsystems are used to verify the controller.