| With the emergence of technologies in the network field,such as the big data processing technology and the cloud computing technology,it further promotes the development of cyber-physical systems(CPSs).However,there are many problems and challenges as well as convenience brought by the wireless network,such as network attacks,packet dropout or disorder,network-induced delay.These problems will cause that the information transmitted to the controller is incomplete and even can not be received,which may destabilize the system.In addition,network communication resources are limited,and uninterrupted transmission will lead to the waste of the communication resources.Therefore,there are lots of studies focusing on the problem of the safe network transmission while saving resources.Under the above research background,this paper designs attack compensation control schemes respectively for model-free adaptive control tracking problems with consideration of different types of malicious network attacks.The main research contents of this paper are as follows:(1)For nonlinear systems with denial-of-service(Do S)attacks and round-trip time(RTT)delays,this study investigates the dynamic event-triggered(DET)model-free adaptive tracking control problem for nonlinear systems.Firstly,we introduce an attack compensation mechanism for the sake of reducing the effects of Do S attacks.At the same time,a network predictive delay compensation mechanism is proposed for compensating the effects of RTT delays caused by packet dropouts,network time-delay and packet disorders.What’s more,the DET method is developed to save communication resources.The nonlinear system is represented by an equivalent data model where the nonlinear characteristics are expressed as pseudo-partial derivative parameters.The designed DET model-free adaptive tracking control algorithm only uses input and output data.(2)For nonlinear systems with Do S attacks,a new input-output constrained(IOC)model-free adaptive tracking control problem for nonlinear constrained systems is studied.Firstly,considering malicious attacks in the network,an attack compensation mechanism is introduced to compensate for Do S attacks.And the IOC strategy is put forward to guarantee the equipment security.Then,for the nonlinear system with unknown model,the new IOC model-free adaptive tracking control algorithm can solve the safe tracking control problem,based on the input and output data.(3)For the subway train system with aperiodic Do S attacks,this section investigates the dynamic event-triggered model-free adaptive security tracking control problem of the subway system with speed and traction/braking force constraints.The complex subway train system is transformed into a linearization data model by using the dynamic linearization technique.Considering aperiodic Do S attacks in the network channel,the attack compensation mechanism is designed.At the same time,for the subway system,both speed and traction/braking forces need to be restrained to ensure the safety of the subway equipment.In addition,the DET mechanism is designed to reduce communication frequency and prolong the hardware life cycle. |
