Dynamic Event-Triggered Model Predictive Control For Networked Control Systems Under DoS Attacks

With the rapid development of modern control technology and network communication technology,low-cost and easy-maintain networked control systems are widely applied in many fields.MPC has become one of the main control methods for handling networked control systems due to its advantages of handling multivariable coupling and optimizing control commands while satisfying system constraints.It is worth noting that,the networked MPC strategies need to consider network security,network bandwidth constraints,and the complexity of MPC computing.Therefore,it is important to reduce the number of signal transmissions in networked MPC,reduce the number of times to solve MPC optimization problem,and effectively responding to cyber attacks.This paper mainly studies the design of dynamic event-triggered MPCs for networked control systems with dual-rate sampled data under bilateral(sensor-controller,controller-actuator)DoS attacks.The specific research content is as follows:(1)An output feedback MPC design method based on dual-channel dynamic eventtriggered transmission mechanisms is proposed for dual-rate sampled data networked control systems with bilateral DoS attacks and time-varying delays.It is assumed that the system output is sampled by two types of sensors with different periods and then transmitted through their channels to the observer at the other end of the network respectively.Firstly,considering the coupling effect of DoS attacks and event-triggered transmissions on networked systems,a kind of attack-dependent dynamic event-triggered transmission mechanism is designed,and different transmission rules are designed based on different scenarios of bilateral DoS attacks to reduce the communication burden caused by frequent triggering.Secondly,a co-design method of dual-rate sampling observer and networked MPC is proposed to reduce the impact of bilateral DoS attacks on the system.By using the estimated state and predicted state of the system to fuse the triggered dual-rate sampled system output,a new matching mechanism to compensate for asynchronous data is proposed to optimize the observer design.Finally,the sufficient conditions for the recursive feasibility of the output feedback MPC algorithm are obtained by using the invariant set principle,and the sufficient conditions for ensuring the performance of the attack-dependent dynamic event-triggered MPC system are obtained by using the Lyapunov-Krasovskii function method.Numerical examples are conducted to verify the effectiveness of the proposed dynamic event-triggered MPC strategy.(2)A dynamic event-triggered dual-mode MPC design method is proposed for nonlinear perturbed networked control systems with bilateral DoS attacks.Firstly,a dualmode MPC architecture,which consists of a local controller,a remote controller,and an intelligent judger that switches between them.The intelligent judger determines the switching between the two controllers through a pre-designed rule.Secondly,considering the impact of the discrete event-triggered transmission mechanism in sensor-controller channels and the DoS attacks on nonlinear networked control systems,a new mathematical model is constructed to describe the presence or absence of attacks,and establish a closedloop system with dual-mode switching controllers.The sufficient conditions for input-tostate practical stability of a dynamic event-triggered dual-mode MPC system are given,and the local controller gain is obtained by using the pole placement method.Finally,a numerical example shows the effectiveness of the proposed dynamic event-triggered dualmode MPC strategy.