Adaptive Control Algorithms For Two Classes Of Nonlinear Cyber-physics Systems Under Deception Attacks

With the development and integration of computer science,artificial intelligence and automation technology,nonlinear cyber-physics systems(CPSs)composed of network and physical components have been widely used in practice,which makes its research of great practical and theoretical significance.On the other hand,adaptive control of nonlinear systems has become a research focus in the field of control.Especially,how to construct effective adaptive control strategies for nonlinear CPSs has attracted many scholars’ research interest.Due to the high openness of the communication network,nonlinear CPSs inevitably have some security risks,such as denial of service attacks,deception attacks,replay attacks,etc.Therefore,how to eliminate the impact of deception attacks and ensure the stability of the nonlinear CPSs is a very challenging topic.This thesis studies two classes of nonlinear CPSs under deception attacks,by proposing a novel coordinate transformation technology and using the attacked states to design a controller,the adaptive control algorithms are constructed to realize the security control of nonlinear CPSs.The specific works are as follows:(1)Two new adaptive fault-tolerant security control schemes are proposed for a class of nonlinear strict feedback CPSs with deception attacks and actuator fault.First of all,the deception attacks change the authenticity of the system states,making it is very hard to realize the tracking control of nonlinear CPSs under deception attacks,and the relevant results have not been proposed.Therefore,an adaptive fault-tolerant tracking control algorithm is designed to defend against deception attacks.By designing a novel coordinate transformation that takes the reference signal and attack gains into account at the same time,the tracking control and the stability of the system are ensured.Nussbaum technology is introduced to eliminate the influence of deception attack on nonlinear CPSs and release the restriction of attack gain symbol.Through the backstepping process,all signals in the closed-loop system are bounded,and the output of the system can track the desired reference signal.Secondly,a more general assumption is proposed for the deception attacks,and a new adaptive fault-tolerant security control scheme is designed to ensure the stability of the attacked system and the boundedness of all signals in the closed-loop system.Finally,the effectiveness of the proposed two algorithms is verified by simulation experiments.(2)An adaptive control scheme based on event-triggered strategy is proposed for a class of non-strict feedback nonlinear CPSs with deception attacks,actuator fault and time-varying delay.First of all,the reasonable assumptions for deception attacks are introduced,and an improved coordinate transformation method is presented to eliminate the difficulties brought by deception attacks on control design.By designing adaptive laws in the process of backstepping,the influence of actuator fault is effectively solved.Secondly,Lyapunov-Krasovskii functions and neural network approximation technology are used to compensate the unknown delay term and deal with the uncertainty of modeling.Further,the relative threshold event-triggered strategy is introduced,and a new event-triggered controller is constructed by using the attacked states to ensure that communication resource will be transmitted only when the trigger conditions are met,thus saving the network bandwidth.The theoretical results show that the proposed control strategy can ensure that the signals are semi-globally bounded in the closed-loop system.Finally,the simulation results show the effectiveness and practicability of the proposed strategy.