Research On Optimal Control Of Networked Control System Under DoS Attack

Networked control system combines the original control system with communication networks.Therefore,networked control system has many characteristics,such as simple wiring,low installation cost,easy maintenance,shared resources,strong mobility and interoperability.It is widely used in the actual industrial control system in various fields,becoming the inevitable trend to the development of control system.However,due to the existence of communication networks,the closure of original control system is destroyed.It will make control system be vulnerable to malicious attacks,which have a serious impact on the normal operation of control system.There are many kinds of malicious attacks against communication networks,among which the most vulnerable is denial-of-service attack.In this paper,networked control system is modeled in the case of denial-of-service attack.Attack strategies and defense strategies are designed by game theory.Optimal controllers are also designed to ensure that networked control system operates stably and achieves optimal performances when suffering denial-of-service attack.The main research results are as follows:Firstly,an optimal control problem is investigated for a networked control system with actuator saturation under denial-of-service attack.Game theory is used to describe interaction between denial-of-service attack and intrusion detection systems introduced into communication networks.Existence and uniqueness of Nash equilibrium strategies are given,as well as an optimal delivery rate is obtained.Under the optimal delivery rate,an iterative adaptive dynamic programming algorithm is applied to get the optimal controller and optimal performance index of the networked control system with actuator saturation.Secondly,an optimal control research on a distributed networked control system is given under denial-of-service attack.Each subsystem is subject to actuator saturation,measurement noise,stochastic disturbances and coupling probabilistic constraints.Then coupling probability constraints are transformed into deterministic constraints.A defense strategy is designed to reduce the impact of denial-of-service attack on the distributed networked control system.The optimal controller is designed to ensure stable operation of the distributed networked control system when all constraints are satisfied.Thirdly,analysis and research on a security problem for a wireless networked control system.Signal-to-noise ratio is introduced and used as a performance index for denial-of-service attackers.Then an optimal attack strategy is designed by cooperative game,which makes the effect of denial-of-service attacks on the wireless networked control system maximized.Furthermore,an optimal controller is designed in the worst case of attacks.It is guaranteed that the wireless networked control system operates stably and achieves optimal performances under the maximum attack.Finally,a networked experimental platform of a linear inverted pendulum is set up.The NetCon system and linear inverted pendulum system are introduced.Then the linear inverted pendulum system is modeled.In the case of denial-of-service attacks,linear quadratic regulator controller not only makes the system stable,but also guarantees the optimal performance of the system.The network experiment platform of the linear inverted pendulum is used to verify it.