Research On Data-driven Control Of Nonlinear Network Systems Under Malicious Attack

In the networked control system,the existence of network channels often causes some unavoidable network problems,such as the malicious attacks,the packet disorder,and the packet loss.Under the influence of the above-mentioned network problems,the control effect of the networked control system will decrease or even destroy the system performance.This thesis considers the output tracking control problem of the nonlinear networked control system under the network attack,the round-trip time(RTT)delay,and the sensor fault.A model-free adaptive control method is used to linearize the system dynamic,and a corresponding compensation mechanism is proposed for the corresponding network problem,so that the system can achieve the desired tracking performance.The specific research contents are as follows:Firstly,considering the RTT delay and denial of service(Do S)attacks in the networked control system,the model-free adaptive security tracking control problem for the nonlinear system is discussed.To compensate the impact of Do S attacks and the RTT delay on the system,a new attack compensation mechanism based on the estimated value and network-based predictive delay compensation mechanism is proposed,respectively.The proposed model-free adaptive security tracking control algorithm can make sure that the output tracking error is bounded in the sense of mean square.An example with comparisons is used to show the effectiveness of the new algorithm.Secondly,the fault-tolerant model-free adaptive control(FT-MFAC)problem of a class of single-input single-output nonlinear networked control systems under Do S attacks is studied.A novel FT-MFAC framework is established with the consideration of Do S attacks and the sensor fault.Based on the designed control framework,an algorithm using only the system output/control input data is proposed,and the effectiveness of the designed algorithm is demonstrated by a simulation.Finally,considering the limitation of network resources,an event-triggered mechanism is designed to save network resources.Based on this mechanism,an eventtriggered model-free adaptive control algorithm is obtained,which is used to study the output tracking problem of the system under the sensor fault and Do S attacks.An example with comparison is given to demonstrate the validity of the new algorithm.