Analysis And Design Of Networked Control Systems With DoS Attack

The networked control system is a multi-dimensional intelligent complex system integrating computing,network and physical environment.Through the organic integration and deep cooperation of computing,communication and control technologies,real-time sensing,dynamic control and information services of large industrial systems are realized.However,with the introduction of the industrial Internet and the networked interaction between human and physical processes,the closure of the original infrastructure system has been broken,and the networked control system is threatened by cyber attacks.Because the cost of Deny-of-Service(DoS)Attack is low,and the negative impact is very large,this thesis mainly studies the security control of networked dynamic systems under DoS attacks.Firstly,for the centralized networked control system,this thesis studies its stabilization control and guaranteed cost control under DoS attack.The networked control system is modeled as a switching control system with a finite number of subsystems,assuming that the maximum number of DoS attacks that the network can withstand is bounded.According to the degree of attack strength,it is divided into three categories: unattack,weak attack and strong attack.Based on the average dwell time method,the sufficient conditions for guaranteeing the stability of the closed-loop system and maintaining the performance level are derived.The convex optimization problem is obtained by the corresponding state feedback controller.Finally,the relationship between the DoS attack frequency and system performance index is established.Further,this thesis gives the upper bound of the frequency that the networked control system can tolerate to be attacked.Secondly,for the multi-agent networked control system,this thesis studies the cases that when a DoS attack occurs,all communication will be invalid,and when there is no attack,the system can maintain the original communication connection.Inthis scenario,the system will present a "switching" feature.On the other hand,due to the communication constraint problem,the sensor adopts the non-uniform sampling mode in the case of keeping the clock synchronization.In this thesis,the network attack model is established by the switching system method,and the non-uniform sampling problem is handled by the input time-delay system method.The closed-loop control system is modeled as a time-delay switching system,and the corresponding distributed state feedback controller and distributed static output feedback controller are designed to achieve consensus goal.Furthermore,the quantitative relationships among sampling period,attack frequency and system performance are analyzed,and the result is applied to the secure cooperative control system design of networked vehicles.Finally,this thesis proposes a distributed secure collaborative control method for multi-agent systems with event-driven communication protocol and DoS attacks.By constructing the event-driven function and combining the switching system method,the time-delay switching system model is obtained,which can be used to describe the characteristics of the closed-loop system.Then,with the average dwell time analysis method,the sufficient conditions for the exponential stability of the closed-loop system are given.Based on some matrix transformation technique,it is shown that the solution of the state feedback controller can be transformed into a feasibility problem of a set of matrix inequalities.Some examples are simulated by Matlab software to verify the effectiveness of the proposed design method.Finally,the thesis is summarized and the prospects for further research are presented.